diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,72660 @@
+{
+  "best_metric": 0.8714,
+  "best_model_checkpoint": "checkpoint/swin-tiny/checkpoint-30969",
+  "epoch": 300.0,
+  "eval_steps": 500,
+  "global_step": 99900,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.0367608070373535,
+      "learning_rate": 9.998998998999e-06,
+      "loss": 4.6422,
+      "step": 10
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.677215099334717,
+      "learning_rate": 9.997997997998e-06,
+      "loss": 4.5999,
+      "step": 20
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.010698318481445,
+      "learning_rate": 9.996996996996998e-06,
+      "loss": 4.5735,
+      "step": 30
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1278815269470215,
+      "learning_rate": 9.995995995995997e-06,
+      "loss": 4.5379,
+      "step": 40
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.357997894287109,
+      "learning_rate": 9.994994994994995e-06,
+      "loss": 4.5355,
+      "step": 50
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.672063827514648,
+      "learning_rate": 9.993993993993994e-06,
+      "loss": 4.5144,
+      "step": 60
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.977757930755615,
+      "learning_rate": 9.992992992992994e-06,
+      "loss": 4.4731,
+      "step": 70
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.93468713760376,
+      "learning_rate": 9.991991991991993e-06,
+      "loss": 4.4178,
+      "step": 80
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.7341413497924805,
+      "learning_rate": 9.990990990990992e-06,
+      "loss": 4.4057,
+      "step": 90
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.587381362915039,
+      "learning_rate": 9.989989989989992e-06,
+      "loss": 4.3499,
+      "step": 100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.376825332641602,
+      "learning_rate": 9.98898898898899e-06,
+      "loss": 4.3059,
+      "step": 110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.455249786376953,
+      "learning_rate": 9.987987987987989e-06,
+      "loss": 4.262,
+      "step": 120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.391611099243164,
+      "learning_rate": 9.986986986986988e-06,
+      "loss": 4.1585,
+      "step": 130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.762007713317871,
+      "learning_rate": 9.985985985985986e-06,
+      "loss": 4.0906,
+      "step": 140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.533121109008789,
+      "learning_rate": 9.984984984984985e-06,
+      "loss": 4.0148,
+      "step": 150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.88967514038086,
+      "learning_rate": 9.983983983983985e-06,
+      "loss": 3.9155,
+      "step": 160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.256155967712402,
+      "learning_rate": 9.982982982982984e-06,
+      "loss": 3.8407,
+      "step": 170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.568601608276367,
+      "learning_rate": 9.981981981981982e-06,
+      "loss": 3.7481,
+      "step": 180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.319816589355469,
+      "learning_rate": 9.980980980980983e-06,
+      "loss": 3.6948,
+      "step": 190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.403288841247559,
+      "learning_rate": 9.979979979979981e-06,
+      "loss": 3.6057,
+      "step": 200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.109684944152832,
+      "learning_rate": 9.97897897897898e-06,
+      "loss": 3.5081,
+      "step": 210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.084592819213867,
+      "learning_rate": 9.977977977977978e-06,
+      "loss": 3.4799,
+      "step": 220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.704780578613281,
+      "learning_rate": 9.976976976976977e-06,
+      "loss": 3.4245,
+      "step": 230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.129402160644531,
+      "learning_rate": 9.975975975975977e-06,
+      "loss": 3.3685,
+      "step": 240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.837770462036133,
+      "learning_rate": 9.974974974974976e-06,
+      "loss": 3.2695,
+      "step": 250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.406055450439453,
+      "learning_rate": 9.973973973973974e-06,
+      "loss": 3.2005,
+      "step": 260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.758235931396484,
+      "learning_rate": 9.972972972972975e-06,
+      "loss": 3.1784,
+      "step": 270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.266830444335938,
+      "learning_rate": 9.971971971971973e-06,
+      "loss": 3.1205,
+      "step": 280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.541472434997559,
+      "learning_rate": 9.970970970970972e-06,
+      "loss": 3.0157,
+      "step": 290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.052526473999023,
+      "learning_rate": 9.96996996996997e-06,
+      "loss": 2.9835,
+      "step": 300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.945847511291504,
+      "learning_rate": 9.968968968968969e-06,
+      "loss": 2.9316,
+      "step": 310
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.683821678161621,
+      "learning_rate": 9.96796796796797e-06,
+      "loss": 2.9266,
+      "step": 320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.834957122802734,
+      "learning_rate": 9.966966966966968e-06,
+      "loss": 2.8188,
+      "step": 330
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.4372,
+      "eval_loss": 2.4231927394866943,
+      "eval_runtime": 15.1879,
+      "eval_samples_per_second": 658.419,
+      "eval_steps_per_second": 2.634,
+      "step": 333
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.244836807250977,
+      "learning_rate": 9.965965965965967e-06,
+      "loss": 2.7959,
+      "step": 340
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.228090286254883,
+      "learning_rate": 9.964964964964965e-06,
+      "loss": 2.7708,
+      "step": 350
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.202799797058105,
+      "learning_rate": 9.963963963963965e-06,
+      "loss": 2.7653,
+      "step": 360
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 24.322519302368164,
+      "learning_rate": 9.962962962962964e-06,
+      "loss": 2.7451,
+      "step": 370
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.843741416931152,
+      "learning_rate": 9.961961961961963e-06,
+      "loss": 2.6411,
+      "step": 380
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.800859451293945,
+      "learning_rate": 9.960960960960961e-06,
+      "loss": 2.6024,
+      "step": 390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.714691162109375,
+      "learning_rate": 9.95995995995996e-06,
+      "loss": 2.5992,
+      "step": 400
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.235062599182129,
+      "learning_rate": 9.95895895895896e-06,
+      "loss": 2.5759,
+      "step": 410
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.454328536987305,
+      "learning_rate": 9.957957957957959e-06,
+      "loss": 2.5312,
+      "step": 420
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.53082275390625,
+      "learning_rate": 9.956956956956957e-06,
+      "loss": 2.4982,
+      "step": 430
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.531620979309082,
+      "learning_rate": 9.955955955955958e-06,
+      "loss": 2.4553,
+      "step": 440
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.312143325805664,
+      "learning_rate": 9.954954954954956e-06,
+      "loss": 2.4303,
+      "step": 450
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.15007209777832,
+      "learning_rate": 9.953953953953955e-06,
+      "loss": 2.4106,
+      "step": 460
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.596813201904297,
+      "learning_rate": 9.952952952952953e-06,
+      "loss": 2.3459,
+      "step": 470
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.045530319213867,
+      "learning_rate": 9.951951951951952e-06,
+      "loss": 2.3336,
+      "step": 480
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.40843391418457,
+      "learning_rate": 9.950950950950952e-06,
+      "loss": 2.2681,
+      "step": 490
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.60472297668457,
+      "learning_rate": 9.949949949949951e-06,
+      "loss": 2.2444,
+      "step": 500
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.754247665405273,
+      "learning_rate": 9.94894894894895e-06,
+      "loss": 2.2344,
+      "step": 510
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.31540870666504,
+      "learning_rate": 9.94794794794795e-06,
+      "loss": 2.3322,
+      "step": 520
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.430391311645508,
+      "learning_rate": 9.946946946946948e-06,
+      "loss": 2.1679,
+      "step": 530
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.180327415466309,
+      "learning_rate": 9.945945945945947e-06,
+      "loss": 2.2569,
+      "step": 540
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 16.49898338317871,
+      "learning_rate": 9.944944944944946e-06,
+      "loss": 2.1861,
+      "step": 550
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.13155460357666,
+      "learning_rate": 9.943943943943944e-06,
+      "loss": 2.136,
+      "step": 560
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.498371124267578,
+      "learning_rate": 9.942942942942944e-06,
+      "loss": 2.195,
+      "step": 570
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 29.57242202758789,
+      "learning_rate": 9.941941941941943e-06,
+      "loss": 2.1519,
+      "step": 580
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.424434661865234,
+      "learning_rate": 9.940940940940942e-06,
+      "loss": 2.1399,
+      "step": 590
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.38271713256836,
+      "learning_rate": 9.93993993993994e-06,
+      "loss": 2.1748,
+      "step": 600
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.109564781188965,
+      "learning_rate": 9.93893893893894e-06,
+      "loss": 2.101,
+      "step": 610
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 19.079050064086914,
+      "learning_rate": 9.937937937937939e-06,
+      "loss": 2.0935,
+      "step": 620
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.897712707519531,
+      "learning_rate": 9.936936936936938e-06,
+      "loss": 2.0537,
+      "step": 630
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.689188003540039,
+      "learning_rate": 9.935935935935936e-06,
+      "loss": 2.0217,
+      "step": 640
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.665223121643066,
+      "learning_rate": 9.934934934934935e-06,
+      "loss": 2.0426,
+      "step": 650
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.258056640625,
+      "learning_rate": 9.933933933933935e-06,
+      "loss": 2.0411,
+      "step": 660
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6269,
+      "eval_loss": 1.4235050678253174,
+      "eval_runtime": 12.7355,
+      "eval_samples_per_second": 785.208,
+      "eval_steps_per_second": 3.141,
+      "step": 666
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.366247177124023,
+      "learning_rate": 9.932932932932934e-06,
+      "loss": 1.9882,
+      "step": 670
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.89084815979004,
+      "learning_rate": 9.931931931931932e-06,
+      "loss": 1.9636,
+      "step": 680
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.520702362060547,
+      "learning_rate": 9.930930930930933e-06,
+      "loss": 1.909,
+      "step": 690
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.264108657836914,
+      "learning_rate": 9.929929929929931e-06,
+      "loss": 1.9156,
+      "step": 700
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.999436378479004,
+      "learning_rate": 9.92892892892893e-06,
+      "loss": 1.9296,
+      "step": 710
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.601841926574707,
+      "learning_rate": 9.927927927927928e-06,
+      "loss": 1.916,
+      "step": 720
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 16.44266128540039,
+      "learning_rate": 9.926926926926927e-06,
+      "loss": 1.9489,
+      "step": 730
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.974905014038086,
+      "learning_rate": 9.925925925925927e-06,
+      "loss": 1.8826,
+      "step": 740
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.924214363098145,
+      "learning_rate": 9.924924924924926e-06,
+      "loss": 1.8674,
+      "step": 750
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.93298053741455,
+      "learning_rate": 9.923923923923925e-06,
+      "loss": 1.8513,
+      "step": 760
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.00749397277832,
+      "learning_rate": 9.922922922922925e-06,
+      "loss": 1.8428,
+      "step": 770
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.950767517089844,
+      "learning_rate": 9.921921921921923e-06,
+      "loss": 1.7832,
+      "step": 780
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.769499778747559,
+      "learning_rate": 9.920920920920922e-06,
+      "loss": 1.7948,
+      "step": 790
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.434747695922852,
+      "learning_rate": 9.91991991991992e-06,
+      "loss": 1.8534,
+      "step": 800
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.389707565307617,
+      "learning_rate": 9.91891891891892e-06,
+      "loss": 1.8856,
+      "step": 810
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.550718307495117,
+      "learning_rate": 9.917917917917918e-06,
+      "loss": 1.8047,
+      "step": 820
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.427532196044922,
+      "learning_rate": 9.916916916916918e-06,
+      "loss": 1.7848,
+      "step": 830
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.922759056091309,
+      "learning_rate": 9.915915915915917e-06,
+      "loss": 1.8626,
+      "step": 840
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 16.29351234436035,
+      "learning_rate": 9.914914914914915e-06,
+      "loss": 1.7767,
+      "step": 850
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.79395866394043,
+      "learning_rate": 9.913913913913916e-06,
+      "loss": 1.665,
+      "step": 860
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.282559394836426,
+      "learning_rate": 9.912912912912914e-06,
+      "loss": 1.7964,
+      "step": 870
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.09691333770752,
+      "learning_rate": 9.911911911911913e-06,
+      "loss": 1.8192,
+      "step": 880
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.11885929107666,
+      "learning_rate": 9.910910910910911e-06,
+      "loss": 1.7374,
+      "step": 890
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.295997619628906,
+      "learning_rate": 9.90990990990991e-06,
+      "loss": 1.7322,
+      "step": 900
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.611212730407715,
+      "learning_rate": 9.90890890890891e-06,
+      "loss": 1.8141,
+      "step": 910
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 15.991599082946777,
+      "learning_rate": 9.907907907907909e-06,
+      "loss": 1.6688,
+      "step": 920
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 15.549467086791992,
+      "learning_rate": 9.906906906906907e-06,
+      "loss": 1.7381,
+      "step": 930
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 14.182323455810547,
+      "learning_rate": 9.905905905905908e-06,
+      "loss": 1.6536,
+      "step": 940
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 14.629528045654297,
+      "learning_rate": 9.904904904904906e-06,
+      "loss": 1.6791,
+      "step": 950
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 17.533554077148438,
+      "learning_rate": 9.903903903903905e-06,
+      "loss": 1.6827,
+      "step": 960
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 15.777318954467773,
+      "learning_rate": 9.902902902902903e-06,
+      "loss": 1.6448,
+      "step": 970
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 18.05567741394043,
+      "learning_rate": 9.901901901901902e-06,
+      "loss": 1.6161,
+      "step": 980
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 19.055246353149414,
+      "learning_rate": 9.900900900900902e-06,
+      "loss": 1.7069,
+      "step": 990
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7102,
+      "eval_loss": 1.055800437927246,
+      "eval_runtime": 12.7654,
+      "eval_samples_per_second": 783.37,
+      "eval_steps_per_second": 3.133,
+      "step": 999
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 17.5743408203125,
+      "learning_rate": 9.899899899899901e-06,
+      "loss": 1.5756,
+      "step": 1000
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 15.102668762207031,
+      "learning_rate": 9.8988988988989e-06,
+      "loss": 1.5809,
+      "step": 1010
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 20.911672592163086,
+      "learning_rate": 9.8978978978979e-06,
+      "loss": 1.6458,
+      "step": 1020
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 13.850802421569824,
+      "learning_rate": 9.896896896896898e-06,
+      "loss": 1.602,
+      "step": 1030
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 15.126977920532227,
+      "learning_rate": 9.895895895895895e-06,
+      "loss": 1.5504,
+      "step": 1040
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 16.3144474029541,
+      "learning_rate": 9.894894894894896e-06,
+      "loss": 1.6521,
+      "step": 1050
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 14.613848686218262,
+      "learning_rate": 9.893893893893894e-06,
+      "loss": 1.5786,
+      "step": 1060
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 15.141481399536133,
+      "learning_rate": 9.892892892892893e-06,
+      "loss": 1.6004,
+      "step": 1070
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 18.73663902282715,
+      "learning_rate": 9.891891891891893e-06,
+      "loss": 1.6059,
+      "step": 1080
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 15.25769329071045,
+      "learning_rate": 9.890890890890892e-06,
+      "loss": 1.634,
+      "step": 1090
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 16.038311004638672,
+      "learning_rate": 9.88988988988989e-06,
+      "loss": 1.5777,
+      "step": 1100
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 16.076766967773438,
+      "learning_rate": 9.88888888888889e-06,
+      "loss": 1.6154,
+      "step": 1110
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 21.815214157104492,
+      "learning_rate": 9.88788788788789e-06,
+      "loss": 1.5889,
+      "step": 1120
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 19.274152755737305,
+      "learning_rate": 9.886886886886888e-06,
+      "loss": 1.5948,
+      "step": 1130
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 14.293699264526367,
+      "learning_rate": 9.885885885885886e-06,
+      "loss": 1.5951,
+      "step": 1140
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 19.594980239868164,
+      "learning_rate": 9.884884884884885e-06,
+      "loss": 1.6085,
+      "step": 1150
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 19.125072479248047,
+      "learning_rate": 9.883883883883885e-06,
+      "loss": 1.6067,
+      "step": 1160
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 19.758527755737305,
+      "learning_rate": 9.882882882882884e-06,
+      "loss": 1.6021,
+      "step": 1170
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 17.54904556274414,
+      "learning_rate": 9.881881881881882e-06,
+      "loss": 1.6093,
+      "step": 1180
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 15.328018188476562,
+      "learning_rate": 9.880880880880883e-06,
+      "loss": 1.4568,
+      "step": 1190
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 15.049988746643066,
+      "learning_rate": 9.879879879879881e-06,
+      "loss": 1.4653,
+      "step": 1200
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 15.316150665283203,
+      "learning_rate": 9.87887887887888e-06,
+      "loss": 1.5626,
+      "step": 1210
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 15.885323524475098,
+      "learning_rate": 9.877877877877879e-06,
+      "loss": 1.5439,
+      "step": 1220
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 14.563307762145996,
+      "learning_rate": 9.876876876876877e-06,
+      "loss": 1.5633,
+      "step": 1230
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 12.845205307006836,
+      "learning_rate": 9.875875875875877e-06,
+      "loss": 1.5821,
+      "step": 1240
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 15.530064582824707,
+      "learning_rate": 9.874874874874876e-06,
+      "loss": 1.4554,
+      "step": 1250
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 14.50391674041748,
+      "learning_rate": 9.873873873873875e-06,
+      "loss": 1.5637,
+      "step": 1260
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 14.745131492614746,
+      "learning_rate": 9.872872872872873e-06,
+      "loss": 1.5024,
+      "step": 1270
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 15.122355461120605,
+      "learning_rate": 9.871871871871873e-06,
+      "loss": 1.4905,
+      "step": 1280
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 15.616936683654785,
+      "learning_rate": 9.87087087087087e-06,
+      "loss": 1.5544,
+      "step": 1290
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 13.670624732971191,
+      "learning_rate": 9.86986986986987e-06,
+      "loss": 1.4526,
+      "step": 1300
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 12.70984172821045,
+      "learning_rate": 9.86886886886887e-06,
+      "loss": 1.4987,
+      "step": 1310
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 15.029037475585938,
+      "learning_rate": 9.867867867867868e-06,
+      "loss": 1.5155,
+      "step": 1320
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 15.3666410446167,
+      "learning_rate": 9.866866866866868e-06,
+      "loss": 1.5722,
+      "step": 1330
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7504,
+      "eval_loss": 0.8656909465789795,
+      "eval_runtime": 12.8593,
+      "eval_samples_per_second": 777.649,
+      "eval_steps_per_second": 3.111,
+      "step": 1332
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 20.454769134521484,
+      "learning_rate": 9.865865865865867e-06,
+      "loss": 1.5871,
+      "step": 1340
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 14.027545928955078,
+      "learning_rate": 9.864864864864865e-06,
+      "loss": 1.4042,
+      "step": 1350
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 25.39883804321289,
+      "learning_rate": 9.863863863863866e-06,
+      "loss": 1.4751,
+      "step": 1360
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 16.596513748168945,
+      "learning_rate": 9.862862862862864e-06,
+      "loss": 1.4249,
+      "step": 1370
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 14.449197769165039,
+      "learning_rate": 9.861861861861863e-06,
+      "loss": 1.492,
+      "step": 1380
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 18.086870193481445,
+      "learning_rate": 9.860860860860861e-06,
+      "loss": 1.4249,
+      "step": 1390
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 18.966970443725586,
+      "learning_rate": 9.85985985985986e-06,
+      "loss": 1.4132,
+      "step": 1400
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 19.73224639892578,
+      "learning_rate": 9.85885885885886e-06,
+      "loss": 1.5375,
+      "step": 1410
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 20.53548812866211,
+      "learning_rate": 9.857857857857859e-06,
+      "loss": 1.4464,
+      "step": 1420
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 16.314666748046875,
+      "learning_rate": 9.856856856856857e-06,
+      "loss": 1.4919,
+      "step": 1430
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 14.961431503295898,
+      "learning_rate": 9.855855855855858e-06,
+      "loss": 1.3864,
+      "step": 1440
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 19.140153884887695,
+      "learning_rate": 9.854854854854856e-06,
+      "loss": 1.3799,
+      "step": 1450
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 15.397107124328613,
+      "learning_rate": 9.853853853853855e-06,
+      "loss": 1.4314,
+      "step": 1460
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 17.417640686035156,
+      "learning_rate": 9.852852852852854e-06,
+      "loss": 1.4163,
+      "step": 1470
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 17.78976058959961,
+      "learning_rate": 9.851851851851852e-06,
+      "loss": 1.3148,
+      "step": 1480
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 15.137016296386719,
+      "learning_rate": 9.85085085085085e-06,
+      "loss": 1.3684,
+      "step": 1490
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 18.988313674926758,
+      "learning_rate": 9.849849849849851e-06,
+      "loss": 1.4638,
+      "step": 1500
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 21.507238388061523,
+      "learning_rate": 9.84884884884885e-06,
+      "loss": 1.4034,
+      "step": 1510
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 14.309231758117676,
+      "learning_rate": 9.847847847847848e-06,
+      "loss": 1.2893,
+      "step": 1520
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 17.174428939819336,
+      "learning_rate": 9.846846846846849e-06,
+      "loss": 1.4635,
+      "step": 1530
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 18.358692169189453,
+      "learning_rate": 9.845845845845845e-06,
+      "loss": 1.3765,
+      "step": 1540
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 17.51491928100586,
+      "learning_rate": 9.844844844844846e-06,
+      "loss": 1.4361,
+      "step": 1550
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 15.559100151062012,
+      "learning_rate": 9.843843843843844e-06,
+      "loss": 1.299,
+      "step": 1560
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 17.001848220825195,
+      "learning_rate": 9.842842842842843e-06,
+      "loss": 1.3858,
+      "step": 1570
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 17.293275833129883,
+      "learning_rate": 9.841841841841843e-06,
+      "loss": 1.3953,
+      "step": 1580
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 20.24597930908203,
+      "learning_rate": 9.840840840840842e-06,
+      "loss": 1.4582,
+      "step": 1590
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 15.464946746826172,
+      "learning_rate": 9.83983983983984e-06,
+      "loss": 1.4351,
+      "step": 1600
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 20.658306121826172,
+      "learning_rate": 9.83883883883884e-06,
+      "loss": 1.3934,
+      "step": 1610
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 14.093180656433105,
+      "learning_rate": 9.83783783783784e-06,
+      "loss": 1.4409,
+      "step": 1620
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 17.276779174804688,
+      "learning_rate": 9.836836836836838e-06,
+      "loss": 1.405,
+      "step": 1630
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 20.552637100219727,
+      "learning_rate": 9.835835835835836e-06,
+      "loss": 1.3978,
+      "step": 1640
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 13.823814392089844,
+      "learning_rate": 9.834834834834835e-06,
+      "loss": 1.4429,
+      "step": 1650
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 18.30078125,
+      "learning_rate": 9.833833833833835e-06,
+      "loss": 1.346,
+      "step": 1660
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.7721,
+      "eval_loss": 0.7773870825767517,
+      "eval_runtime": 12.5993,
+      "eval_samples_per_second": 793.697,
+      "eval_steps_per_second": 3.175,
+      "step": 1665
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 13.90429973602295,
+      "learning_rate": 9.832832832832834e-06,
+      "loss": 1.1999,
+      "step": 1670
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 17.547555923461914,
+      "learning_rate": 9.831831831831833e-06,
+      "loss": 1.3139,
+      "step": 1680
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 17.695215225219727,
+      "learning_rate": 9.830830830830833e-06,
+      "loss": 1.3693,
+      "step": 1690
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 13.60401439666748,
+      "learning_rate": 9.829829829829831e-06,
+      "loss": 1.3242,
+      "step": 1700
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 17.991899490356445,
+      "learning_rate": 9.82882882882883e-06,
+      "loss": 1.3636,
+      "step": 1710
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 14.60073471069336,
+      "learning_rate": 9.827827827827829e-06,
+      "loss": 1.2807,
+      "step": 1720
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 16.130075454711914,
+      "learning_rate": 9.826826826826827e-06,
+      "loss": 1.4117,
+      "step": 1730
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 24.69490623474121,
+      "learning_rate": 9.825825825825826e-06,
+      "loss": 1.3732,
+      "step": 1740
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 14.710121154785156,
+      "learning_rate": 9.824824824824826e-06,
+      "loss": 1.2513,
+      "step": 1750
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 18.907339096069336,
+      "learning_rate": 9.823823823823825e-06,
+      "loss": 1.3643,
+      "step": 1760
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 14.353303909301758,
+      "learning_rate": 9.822822822822823e-06,
+      "loss": 1.3138,
+      "step": 1770
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 17.80875015258789,
+      "learning_rate": 9.821821821821824e-06,
+      "loss": 1.259,
+      "step": 1780
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 22.915904998779297,
+      "learning_rate": 9.82082082082082e-06,
+      "loss": 1.3137,
+      "step": 1790
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 15.092241287231445,
+      "learning_rate": 9.81981981981982e-06,
+      "loss": 1.3528,
+      "step": 1800
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 19.2230281829834,
+      "learning_rate": 9.81881881881882e-06,
+      "loss": 1.2926,
+      "step": 1810
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 15.056241989135742,
+      "learning_rate": 9.817817817817818e-06,
+      "loss": 1.3463,
+      "step": 1820
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 15.7086181640625,
+      "learning_rate": 9.816816816816818e-06,
+      "loss": 1.3135,
+      "step": 1830
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 21.66438102722168,
+      "learning_rate": 9.815815815815817e-06,
+      "loss": 1.3012,
+      "step": 1840
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 17.32779884338379,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 1.3644,
+      "step": 1850
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 15.512614250183105,
+      "learning_rate": 9.813813813813816e-06,
+      "loss": 1.2463,
+      "step": 1860
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 16.07647705078125,
+      "learning_rate": 9.812812812812814e-06,
+      "loss": 1.3773,
+      "step": 1870
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 13.52031421661377,
+      "learning_rate": 9.811811811811813e-06,
+      "loss": 1.3064,
+      "step": 1880
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 14.820934295654297,
+      "learning_rate": 9.810810810810811e-06,
+      "loss": 1.2048,
+      "step": 1890
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 17.276151657104492,
+      "learning_rate": 9.80980980980981e-06,
+      "loss": 1.3738,
+      "step": 1900
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 13.631850242614746,
+      "learning_rate": 9.80880880880881e-06,
+      "loss": 1.3378,
+      "step": 1910
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 19.11239242553711,
+      "learning_rate": 9.807807807807809e-06,
+      "loss": 1.2402,
+      "step": 1920
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 14.28607177734375,
+      "learning_rate": 9.806806806806808e-06,
+      "loss": 1.3308,
+      "step": 1930
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 13.145687103271484,
+      "learning_rate": 9.805805805805808e-06,
+      "loss": 1.324,
+      "step": 1940
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 15.426340103149414,
+      "learning_rate": 9.804804804804806e-06,
+      "loss": 1.3066,
+      "step": 1950
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 11.67961597442627,
+      "learning_rate": 9.803803803803803e-06,
+      "loss": 1.3375,
+      "step": 1960
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 15.971795082092285,
+      "learning_rate": 9.802802802802804e-06,
+      "loss": 1.2993,
+      "step": 1970
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 16.692211151123047,
+      "learning_rate": 9.801801801801802e-06,
+      "loss": 1.2503,
+      "step": 1980
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 14.695980072021484,
+      "learning_rate": 9.8008008008008e-06,
+      "loss": 1.303,
+      "step": 1990
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.7874,
+      "eval_loss": 0.7137871980667114,
+      "eval_runtime": 12.7077,
+      "eval_samples_per_second": 786.925,
+      "eval_steps_per_second": 3.148,
+      "step": 1998
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 13.036826133728027,
+      "learning_rate": 9.799799799799801e-06,
+      "loss": 1.5263,
+      "step": 2000
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 20.32438087463379,
+      "learning_rate": 9.7987987987988e-06,
+      "loss": 1.3226,
+      "step": 2010
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 17.317169189453125,
+      "learning_rate": 9.797797797797798e-06,
+      "loss": 1.2233,
+      "step": 2020
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 17.734073638916016,
+      "learning_rate": 9.796796796796799e-06,
+      "loss": 1.1739,
+      "step": 2030
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 15.11571216583252,
+      "learning_rate": 9.795795795795795e-06,
+      "loss": 1.2039,
+      "step": 2040
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 17.901887893676758,
+      "learning_rate": 9.794794794794796e-06,
+      "loss": 1.2837,
+      "step": 2050
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 16.518293380737305,
+      "learning_rate": 9.793793793793794e-06,
+      "loss": 1.2622,
+      "step": 2060
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 16.371145248413086,
+      "learning_rate": 9.792792792792793e-06,
+      "loss": 1.2561,
+      "step": 2070
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 13.058977127075195,
+      "learning_rate": 9.791791791791793e-06,
+      "loss": 1.2545,
+      "step": 2080
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 16.877670288085938,
+      "learning_rate": 9.790790790790792e-06,
+      "loss": 1.1889,
+      "step": 2090
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 19.347734451293945,
+      "learning_rate": 9.78978978978979e-06,
+      "loss": 1.2674,
+      "step": 2100
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 16.47791290283203,
+      "learning_rate": 9.78878878878879e-06,
+      "loss": 1.254,
+      "step": 2110
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 24.682647705078125,
+      "learning_rate": 9.787787787787788e-06,
+      "loss": 1.2575,
+      "step": 2120
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 31.932409286499023,
+      "learning_rate": 9.786786786786788e-06,
+      "loss": 1.3132,
+      "step": 2130
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 12.494312286376953,
+      "learning_rate": 9.785785785785787e-06,
+      "loss": 1.1676,
+      "step": 2140
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 24.20763397216797,
+      "learning_rate": 9.784784784784785e-06,
+      "loss": 1.3191,
+      "step": 2150
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 14.357993125915527,
+      "learning_rate": 9.783783783783785e-06,
+      "loss": 1.2434,
+      "step": 2160
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 17.260360717773438,
+      "learning_rate": 9.782782782782784e-06,
+      "loss": 1.2659,
+      "step": 2170
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 17.088218688964844,
+      "learning_rate": 9.781781781781783e-06,
+      "loss": 1.2073,
+      "step": 2180
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 17.340211868286133,
+      "learning_rate": 9.780780780780781e-06,
+      "loss": 1.2804,
+      "step": 2190
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 19.425783157348633,
+      "learning_rate": 9.779779779779781e-06,
+      "loss": 1.2694,
+      "step": 2200
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 14.971159934997559,
+      "learning_rate": 9.778778778778778e-06,
+      "loss": 1.2434,
+      "step": 2210
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 15.645931243896484,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 1.309,
+      "step": 2220
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 14.082205772399902,
+      "learning_rate": 9.776776776776777e-06,
+      "loss": 1.2739,
+      "step": 2230
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 14.896344184875488,
+      "learning_rate": 9.775775775775776e-06,
+      "loss": 1.2321,
+      "step": 2240
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 14.076533317565918,
+      "learning_rate": 9.774774774774776e-06,
+      "loss": 1.0925,
+      "step": 2250
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 17.046342849731445,
+      "learning_rate": 9.773773773773775e-06,
+      "loss": 1.2189,
+      "step": 2260
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 23.03464126586914,
+      "learning_rate": 9.772772772772773e-06,
+      "loss": 1.1931,
+      "step": 2270
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 20.006450653076172,
+      "learning_rate": 9.771771771771774e-06,
+      "loss": 1.2965,
+      "step": 2280
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 45.20359420776367,
+      "learning_rate": 9.77077077077077e-06,
+      "loss": 1.2089,
+      "step": 2290
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 24.090991973876953,
+      "learning_rate": 9.76976976976977e-06,
+      "loss": 1.1629,
+      "step": 2300
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 21.395421981811523,
+      "learning_rate": 9.76876876876877e-06,
+      "loss": 1.2583,
+      "step": 2310
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 17.435306549072266,
+      "learning_rate": 9.767767767767768e-06,
+      "loss": 1.2224,
+      "step": 2320
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 14.937963485717773,
+      "learning_rate": 9.766766766766768e-06,
+      "loss": 1.2045,
+      "step": 2330
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7986,
+      "eval_loss": 0.6616324186325073,
+      "eval_runtime": 12.886,
+      "eval_samples_per_second": 776.034,
+      "eval_steps_per_second": 3.104,
+      "step": 2331
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 17.74285125732422,
+      "learning_rate": 9.765765765765767e-06,
+      "loss": 1.213,
+      "step": 2340
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 18.209003448486328,
+      "learning_rate": 9.764764764764765e-06,
+      "loss": 1.2171,
+      "step": 2350
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 16.93661117553711,
+      "learning_rate": 9.763763763763766e-06,
+      "loss": 1.2575,
+      "step": 2360
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 13.612496376037598,
+      "learning_rate": 9.762762762762763e-06,
+      "loss": 1.1219,
+      "step": 2370
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 15.661642074584961,
+      "learning_rate": 9.761761761761763e-06,
+      "loss": 1.1897,
+      "step": 2380
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 14.000067710876465,
+      "learning_rate": 9.760760760760762e-06,
+      "loss": 1.1881,
+      "step": 2390
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 15.062593460083008,
+      "learning_rate": 9.75975975975976e-06,
+      "loss": 1.2742,
+      "step": 2400
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 16.873779296875,
+      "learning_rate": 9.758758758758759e-06,
+      "loss": 1.1432,
+      "step": 2410
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 14.98442554473877,
+      "learning_rate": 9.757757757757759e-06,
+      "loss": 1.1578,
+      "step": 2420
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 20.095666885375977,
+      "learning_rate": 9.756756756756758e-06,
+      "loss": 1.1644,
+      "step": 2430
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 17.67222023010254,
+      "learning_rate": 9.755755755755756e-06,
+      "loss": 1.1308,
+      "step": 2440
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 12.987893104553223,
+      "learning_rate": 9.754754754754756e-06,
+      "loss": 1.218,
+      "step": 2450
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 14.277360916137695,
+      "learning_rate": 9.753753753753753e-06,
+      "loss": 1.1472,
+      "step": 2460
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 20.748971939086914,
+      "learning_rate": 9.752752752752754e-06,
+      "loss": 1.2294,
+      "step": 2470
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 16.136808395385742,
+      "learning_rate": 9.751751751751752e-06,
+      "loss": 1.1492,
+      "step": 2480
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 13.233691215515137,
+      "learning_rate": 9.750750750750751e-06,
+      "loss": 1.186,
+      "step": 2490
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 18.350692749023438,
+      "learning_rate": 9.749749749749751e-06,
+      "loss": 1.2337,
+      "step": 2500
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 18.468725204467773,
+      "learning_rate": 9.74874874874875e-06,
+      "loss": 1.2043,
+      "step": 2510
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 15.029038429260254,
+      "learning_rate": 9.747747747747748e-06,
+      "loss": 1.2252,
+      "step": 2520
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 15.551007270812988,
+      "learning_rate": 9.746746746746749e-06,
+      "loss": 1.1802,
+      "step": 2530
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 13.695267677307129,
+      "learning_rate": 9.745745745745746e-06,
+      "loss": 1.1828,
+      "step": 2540
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 22.22896385192871,
+      "learning_rate": 9.744744744744746e-06,
+      "loss": 1.1379,
+      "step": 2550
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 14.21728229522705,
+      "learning_rate": 9.743743743743744e-06,
+      "loss": 1.1172,
+      "step": 2560
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 19.173032760620117,
+      "learning_rate": 9.742742742742743e-06,
+      "loss": 1.2143,
+      "step": 2570
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 24.064266204833984,
+      "learning_rate": 9.741741741741743e-06,
+      "loss": 1.1707,
+      "step": 2580
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 19.597206115722656,
+      "learning_rate": 9.740740740740742e-06,
+      "loss": 1.1723,
+      "step": 2590
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 15.806962966918945,
+      "learning_rate": 9.73973973973974e-06,
+      "loss": 1.2468,
+      "step": 2600
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 19.199220657348633,
+      "learning_rate": 9.73873873873874e-06,
+      "loss": 1.2025,
+      "step": 2610
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 19.19084358215332,
+      "learning_rate": 9.737737737737738e-06,
+      "loss": 1.1438,
+      "step": 2620
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 14.284010887145996,
+      "learning_rate": 9.736736736736738e-06,
+      "loss": 1.2175,
+      "step": 2630
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 19.574195861816406,
+      "learning_rate": 9.735735735735737e-06,
+      "loss": 1.0988,
+      "step": 2640
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 13.940256118774414,
+      "learning_rate": 9.734734734734735e-06,
+      "loss": 1.1101,
+      "step": 2650
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 20.282869338989258,
+      "learning_rate": 9.733733733733734e-06,
+      "loss": 1.2482,
+      "step": 2660
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8128,
+      "eval_loss": 0.6209900379180908,
+      "eval_runtime": 12.6026,
+      "eval_samples_per_second": 793.49,
+      "eval_steps_per_second": 3.174,
+      "step": 2664
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 16.045846939086914,
+      "learning_rate": 9.732732732732734e-06,
+      "loss": 1.1542,
+      "step": 2670
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 14.654650688171387,
+      "learning_rate": 9.731731731731733e-06,
+      "loss": 1.2251,
+      "step": 2680
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 19.446109771728516,
+      "learning_rate": 9.730730730730731e-06,
+      "loss": 1.1321,
+      "step": 2690
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 14.171721458435059,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 1.1187,
+      "step": 2700
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 18.1235408782959,
+      "learning_rate": 9.728728728728728e-06,
+      "loss": 1.1489,
+      "step": 2710
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 21.12098503112793,
+      "learning_rate": 9.727727727727729e-06,
+      "loss": 1.1572,
+      "step": 2720
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 13.129083633422852,
+      "learning_rate": 9.726726726726727e-06,
+      "loss": 1.173,
+      "step": 2730
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 14.0629301071167,
+      "learning_rate": 9.725725725725726e-06,
+      "loss": 1.157,
+      "step": 2740
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 21.699491500854492,
+      "learning_rate": 9.724724724724726e-06,
+      "loss": 1.1807,
+      "step": 2750
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 15.412301063537598,
+      "learning_rate": 9.723723723723725e-06,
+      "loss": 1.1862,
+      "step": 2760
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 21.340255737304688,
+      "learning_rate": 9.722722722722723e-06,
+      "loss": 1.1649,
+      "step": 2770
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 14.753629684448242,
+      "learning_rate": 9.721721721721724e-06,
+      "loss": 1.156,
+      "step": 2780
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 18.49135398864746,
+      "learning_rate": 9.72072072072072e-06,
+      "loss": 1.1557,
+      "step": 2790
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 16.890783309936523,
+      "learning_rate": 9.719719719719721e-06,
+      "loss": 1.1386,
+      "step": 2800
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 16.26944351196289,
+      "learning_rate": 9.71871871871872e-06,
+      "loss": 1.1508,
+      "step": 2810
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 13.584962844848633,
+      "learning_rate": 9.717717717717718e-06,
+      "loss": 1.0747,
+      "step": 2820
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 19.696990966796875,
+      "learning_rate": 9.716716716716718e-06,
+      "loss": 1.1413,
+      "step": 2830
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 19.515819549560547,
+      "learning_rate": 9.715715715715717e-06,
+      "loss": 1.129,
+      "step": 2840
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 26.190208435058594,
+      "learning_rate": 9.714714714714716e-06,
+      "loss": 1.1476,
+      "step": 2850
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 16.262331008911133,
+      "learning_rate": 9.713713713713714e-06,
+      "loss": 1.1215,
+      "step": 2860
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 14.385309219360352,
+      "learning_rate": 9.712712712712713e-06,
+      "loss": 1.093,
+      "step": 2870
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 18.503665924072266,
+      "learning_rate": 9.711711711711711e-06,
+      "loss": 1.0973,
+      "step": 2880
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 17.09222984313965,
+      "learning_rate": 9.710710710710712e-06,
+      "loss": 1.1493,
+      "step": 2890
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 17.18866729736328,
+      "learning_rate": 9.70970970970971e-06,
+      "loss": 1.1105,
+      "step": 2900
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 12.714863777160645,
+      "learning_rate": 9.708708708708709e-06,
+      "loss": 1.1722,
+      "step": 2910
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 18.195629119873047,
+      "learning_rate": 9.707707707707709e-06,
+      "loss": 1.1124,
+      "step": 2920
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 16.43483543395996,
+      "learning_rate": 9.706706706706708e-06,
+      "loss": 1.0822,
+      "step": 2930
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 20.604055404663086,
+      "learning_rate": 9.705705705705706e-06,
+      "loss": 1.1663,
+      "step": 2940
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 16.31377410888672,
+      "learning_rate": 9.704704704704707e-06,
+      "loss": 1.0176,
+      "step": 2950
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 17.643062591552734,
+      "learning_rate": 9.703703703703703e-06,
+      "loss": 1.208,
+      "step": 2960
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 14.060117721557617,
+      "learning_rate": 9.702702702702704e-06,
+      "loss": 1.1233,
+      "step": 2970
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 19.547208786010742,
+      "learning_rate": 9.701701701701702e-06,
+      "loss": 1.0834,
+      "step": 2980
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 16.685522079467773,
+      "learning_rate": 9.700700700700701e-06,
+      "loss": 1.1202,
+      "step": 2990
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.8185,
+      "eval_loss": 0.5924676060676575,
+      "eval_runtime": 12.456,
+      "eval_samples_per_second": 802.828,
+      "eval_steps_per_second": 3.211,
+      "step": 2997
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 12.231082916259766,
+      "learning_rate": 9.699699699699701e-06,
+      "loss": 1.2633,
+      "step": 3000
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 15.255853652954102,
+      "learning_rate": 9.6986986986987e-06,
+      "loss": 1.0957,
+      "step": 3010
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 14.946885108947754,
+      "learning_rate": 9.697697697697698e-06,
+      "loss": 1.1606,
+      "step": 3020
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 15.070345878601074,
+      "learning_rate": 9.696696696696699e-06,
+      "loss": 1.0868,
+      "step": 3030
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 17.080293655395508,
+      "learning_rate": 9.695695695695696e-06,
+      "loss": 1.0463,
+      "step": 3040
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 17.268754959106445,
+      "learning_rate": 9.694694694694696e-06,
+      "loss": 1.1091,
+      "step": 3050
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 23.904386520385742,
+      "learning_rate": 9.693693693693694e-06,
+      "loss": 1.1861,
+      "step": 3060
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 21.123323440551758,
+      "learning_rate": 9.692692692692693e-06,
+      "loss": 1.0999,
+      "step": 3070
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 13.559005737304688,
+      "learning_rate": 9.691691691691693e-06,
+      "loss": 1.0932,
+      "step": 3080
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 24.972814559936523,
+      "learning_rate": 9.690690690690692e-06,
+      "loss": 0.9839,
+      "step": 3090
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 14.8424654006958,
+      "learning_rate": 9.68968968968969e-06,
+      "loss": 1.1106,
+      "step": 3100
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 15.940447807312012,
+      "learning_rate": 9.68868868868869e-06,
+      "loss": 1.1486,
+      "step": 3110
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 12.488078117370605,
+      "learning_rate": 9.687687687687688e-06,
+      "loss": 1.1388,
+      "step": 3120
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 18.363967895507812,
+      "learning_rate": 9.686686686686686e-06,
+      "loss": 1.1205,
+      "step": 3130
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 15.243896484375,
+      "learning_rate": 9.685685685685687e-06,
+      "loss": 1.1399,
+      "step": 3140
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 15.083489418029785,
+      "learning_rate": 9.684684684684685e-06,
+      "loss": 1.0967,
+      "step": 3150
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 17.603118896484375,
+      "learning_rate": 9.683683683683684e-06,
+      "loss": 1.0997,
+      "step": 3160
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 13.779284477233887,
+      "learning_rate": 9.682682682682684e-06,
+      "loss": 1.0508,
+      "step": 3170
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 17.325523376464844,
+      "learning_rate": 9.681681681681683e-06,
+      "loss": 1.2167,
+      "step": 3180
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 20.735912322998047,
+      "learning_rate": 9.680680680680681e-06,
+      "loss": 1.0901,
+      "step": 3190
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 15.422182083129883,
+      "learning_rate": 9.67967967967968e-06,
+      "loss": 1.0652,
+      "step": 3200
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 14.05704402923584,
+      "learning_rate": 9.678678678678679e-06,
+      "loss": 1.0333,
+      "step": 3210
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 18.3939208984375,
+      "learning_rate": 9.677677677677679e-06,
+      "loss": 1.1638,
+      "step": 3220
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 17.519819259643555,
+      "learning_rate": 9.676676676676677e-06,
+      "loss": 1.0882,
+      "step": 3230
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 14.839508056640625,
+      "learning_rate": 9.675675675675676e-06,
+      "loss": 1.0706,
+      "step": 3240
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 16.574199676513672,
+      "learning_rate": 9.674674674674676e-06,
+      "loss": 1.1276,
+      "step": 3250
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 14.425287246704102,
+      "learning_rate": 9.673673673673675e-06,
+      "loss": 1.0369,
+      "step": 3260
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 11.453363418579102,
+      "learning_rate": 9.672672672672673e-06,
+      "loss": 1.13,
+      "step": 3270
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 15.991726875305176,
+      "learning_rate": 9.671671671671674e-06,
+      "loss": 1.1164,
+      "step": 3280
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 19.66295623779297,
+      "learning_rate": 9.67067067067067e-06,
+      "loss": 1.0496,
+      "step": 3290
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 17.811168670654297,
+      "learning_rate": 9.669669669669671e-06,
+      "loss": 1.0741,
+      "step": 3300
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 16.96430015563965,
+      "learning_rate": 9.66866866866867e-06,
+      "loss": 1.1444,
+      "step": 3310
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 17.067520141601562,
+      "learning_rate": 9.667667667667668e-06,
+      "loss": 1.0883,
+      "step": 3320
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 27.854724884033203,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 1.0021,
+      "step": 3330
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.8235,
+      "eval_loss": 0.5728312730789185,
+      "eval_runtime": 12.787,
+      "eval_samples_per_second": 782.045,
+      "eval_steps_per_second": 3.128,
+      "step": 3330
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 15.601628303527832,
+      "learning_rate": 9.665665665665667e-06,
+      "loss": 1.0575,
+      "step": 3340
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 16.880401611328125,
+      "learning_rate": 9.664664664664666e-06,
+      "loss": 1.0543,
+      "step": 3350
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 16.332054138183594,
+      "learning_rate": 9.663663663663664e-06,
+      "loss": 1.0403,
+      "step": 3360
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 16.27848243713379,
+      "learning_rate": 9.662662662662663e-06,
+      "loss": 1.0386,
+      "step": 3370
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 15.288803100585938,
+      "learning_rate": 9.661661661661661e-06,
+      "loss": 1.1157,
+      "step": 3380
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 19.522502899169922,
+      "learning_rate": 9.660660660660662e-06,
+      "loss": 1.0648,
+      "step": 3390
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 17.908447265625,
+      "learning_rate": 9.65965965965966e-06,
+      "loss": 1.1023,
+      "step": 3400
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 13.076386451721191,
+      "learning_rate": 9.658658658658659e-06,
+      "loss": 1.032,
+      "step": 3410
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 15.524860382080078,
+      "learning_rate": 9.65765765765766e-06,
+      "loss": 1.0363,
+      "step": 3420
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 14.770012855529785,
+      "learning_rate": 9.656656656656658e-06,
+      "loss": 1.0339,
+      "step": 3430
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 19.564733505249023,
+      "learning_rate": 9.655655655655656e-06,
+      "loss": 1.0874,
+      "step": 3440
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 16.416458129882812,
+      "learning_rate": 9.654654654654655e-06,
+      "loss": 1.0911,
+      "step": 3450
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 18.273672103881836,
+      "learning_rate": 9.653653653653654e-06,
+      "loss": 1.0919,
+      "step": 3460
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 15.215808868408203,
+      "learning_rate": 9.652652652652654e-06,
+      "loss": 1.0825,
+      "step": 3470
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 15.298432350158691,
+      "learning_rate": 9.651651651651652e-06,
+      "loss": 1.086,
+      "step": 3480
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 12.385087966918945,
+      "learning_rate": 9.650650650650651e-06,
+      "loss": 1.1275,
+      "step": 3490
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 16.057527542114258,
+      "learning_rate": 9.649649649649651e-06,
+      "loss": 1.0923,
+      "step": 3500
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 15.67593765258789,
+      "learning_rate": 9.64864864864865e-06,
+      "loss": 1.0532,
+      "step": 3510
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 15.688575744628906,
+      "learning_rate": 9.647647647647648e-06,
+      "loss": 1.0284,
+      "step": 3520
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 14.33437442779541,
+      "learning_rate": 9.646646646646649e-06,
+      "loss": 1.057,
+      "step": 3530
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 17.05129623413086,
+      "learning_rate": 9.645645645645646e-06,
+      "loss": 1.0499,
+      "step": 3540
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 17.67121696472168,
+      "learning_rate": 9.644644644644644e-06,
+      "loss": 1.0749,
+      "step": 3550
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 16.625755310058594,
+      "learning_rate": 9.643643643643645e-06,
+      "loss": 1.0579,
+      "step": 3560
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 19.9012508392334,
+      "learning_rate": 9.642642642642643e-06,
+      "loss": 1.0356,
+      "step": 3570
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 21.382139205932617,
+      "learning_rate": 9.641641641641642e-06,
+      "loss": 1.0528,
+      "step": 3580
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 18.916330337524414,
+      "learning_rate": 9.640640640640642e-06,
+      "loss": 0.9962,
+      "step": 3590
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 18.414865493774414,
+      "learning_rate": 9.63963963963964e-06,
+      "loss": 1.0563,
+      "step": 3600
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 13.265142440795898,
+      "learning_rate": 9.63863863863864e-06,
+      "loss": 0.9864,
+      "step": 3610
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 14.302437782287598,
+      "learning_rate": 9.637637637637638e-06,
+      "loss": 1.0679,
+      "step": 3620
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 17.272510528564453,
+      "learning_rate": 9.636636636636636e-06,
+      "loss": 1.0747,
+      "step": 3630
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 17.251792907714844,
+      "learning_rate": 9.635635635635637e-06,
+      "loss": 1.0952,
+      "step": 3640
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 20.940811157226562,
+      "learning_rate": 9.634634634634635e-06,
+      "loss": 1.0542,
+      "step": 3650
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 18.766862869262695,
+      "learning_rate": 9.633633633633634e-06,
+      "loss": 1.0662,
+      "step": 3660
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.829,
+      "eval_loss": 0.5636696815490723,
+      "eval_runtime": 12.5849,
+      "eval_samples_per_second": 794.602,
+      "eval_steps_per_second": 3.178,
+      "step": 3663
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 14.225459098815918,
+      "learning_rate": 9.632632632632634e-06,
+      "loss": 1.2346,
+      "step": 3670
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 13.948782920837402,
+      "learning_rate": 9.631631631631633e-06,
+      "loss": 1.0052,
+      "step": 3680
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 17.608837127685547,
+      "learning_rate": 9.630630630630631e-06,
+      "loss": 1.0115,
+      "step": 3690
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 23.72847557067871,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 1.0427,
+      "step": 3700
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 16.49607276916504,
+      "learning_rate": 9.628628628628629e-06,
+      "loss": 1.0758,
+      "step": 3710
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 15.795092582702637,
+      "learning_rate": 9.627627627627629e-06,
+      "loss": 0.9999,
+      "step": 3720
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 15.525904655456543,
+      "learning_rate": 9.626626626626627e-06,
+      "loss": 1.0969,
+      "step": 3730
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 16.011144638061523,
+      "learning_rate": 9.625625625625626e-06,
+      "loss": 0.9972,
+      "step": 3740
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 17.082075119018555,
+      "learning_rate": 9.624624624624626e-06,
+      "loss": 1.023,
+      "step": 3750
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 17.181434631347656,
+      "learning_rate": 9.623623623623625e-06,
+      "loss": 1.0567,
+      "step": 3760
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 23.32236099243164,
+      "learning_rate": 9.622622622622624e-06,
+      "loss": 1.1356,
+      "step": 3770
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 14.4651460647583,
+      "learning_rate": 9.621621621621622e-06,
+      "loss": 1.1166,
+      "step": 3780
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 16.139591217041016,
+      "learning_rate": 9.62062062062062e-06,
+      "loss": 1.0839,
+      "step": 3790
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 12.324134826660156,
+      "learning_rate": 9.61961961961962e-06,
+      "loss": 1.0817,
+      "step": 3800
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 15.287452697753906,
+      "learning_rate": 9.61861861861862e-06,
+      "loss": 1.0017,
+      "step": 3810
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 18.197134017944336,
+      "learning_rate": 9.617617617617618e-06,
+      "loss": 1.0289,
+      "step": 3820
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 17.314950942993164,
+      "learning_rate": 9.616616616616617e-06,
+      "loss": 1.0783,
+      "step": 3830
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 16.530426025390625,
+      "learning_rate": 9.615615615615617e-06,
+      "loss": 1.0625,
+      "step": 3840
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 18.19109535217285,
+      "learning_rate": 9.614614614614616e-06,
+      "loss": 1.0336,
+      "step": 3850
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 17.461894989013672,
+      "learning_rate": 9.613613613613614e-06,
+      "loss": 1.078,
+      "step": 3860
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 15.058967590332031,
+      "learning_rate": 9.612612612612613e-06,
+      "loss": 1.0271,
+      "step": 3870
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 16.8382568359375,
+      "learning_rate": 9.611611611611611e-06,
+      "loss": 1.0783,
+      "step": 3880
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 16.57802391052246,
+      "learning_rate": 9.610610610610612e-06,
+      "loss": 1.0591,
+      "step": 3890
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 19.649625778198242,
+      "learning_rate": 9.60960960960961e-06,
+      "loss": 1.0703,
+      "step": 3900
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 23.350257873535156,
+      "learning_rate": 9.608608608608609e-06,
+      "loss": 1.1037,
+      "step": 3910
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 28.183189392089844,
+      "learning_rate": 9.60760760760761e-06,
+      "loss": 1.0253,
+      "step": 3920
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 21.10226058959961,
+      "learning_rate": 9.606606606606608e-06,
+      "loss": 1.0825,
+      "step": 3930
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 19.632354736328125,
+      "learning_rate": 9.605605605605606e-06,
+      "loss": 0.9878,
+      "step": 3940
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 17.59662437438965,
+      "learning_rate": 9.604604604604605e-06,
+      "loss": 1.0505,
+      "step": 3950
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 18.002777099609375,
+      "learning_rate": 9.603603603603604e-06,
+      "loss": 1.0303,
+      "step": 3960
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 19.330671310424805,
+      "learning_rate": 9.602602602602604e-06,
+      "loss": 1.0271,
+      "step": 3970
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 14.030105590820312,
+      "learning_rate": 9.601601601601602e-06,
+      "loss": 1.0634,
+      "step": 3980
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 27.249462127685547,
+      "learning_rate": 9.600600600600601e-06,
+      "loss": 1.0263,
+      "step": 3990
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.8303,
+      "eval_loss": 0.5441622734069824,
+      "eval_runtime": 12.8921,
+      "eval_samples_per_second": 775.671,
+      "eval_steps_per_second": 3.103,
+      "step": 3996
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 21.14763641357422,
+      "learning_rate": 9.5995995995996e-06,
+      "loss": 1.209,
+      "step": 4000
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 16.916152954101562,
+      "learning_rate": 9.5985985985986e-06,
+      "loss": 0.964,
+      "step": 4010
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 20.873559951782227,
+      "learning_rate": 9.597597597597599e-06,
+      "loss": 0.9949,
+      "step": 4020
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 14.839821815490723,
+      "learning_rate": 9.596596596596597e-06,
+      "loss": 0.9582,
+      "step": 4030
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 17.228927612304688,
+      "learning_rate": 9.595595595595596e-06,
+      "loss": 1.0007,
+      "step": 4040
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 27.08382797241211,
+      "learning_rate": 9.594594594594594e-06,
+      "loss": 1.0368,
+      "step": 4050
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 15.229191780090332,
+      "learning_rate": 9.593593593593595e-06,
+      "loss": 1.0135,
+      "step": 4060
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 14.073891639709473,
+      "learning_rate": 9.592592592592593e-06,
+      "loss": 1.0114,
+      "step": 4070
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 16.11541748046875,
+      "learning_rate": 9.591591591591592e-06,
+      "loss": 1.0698,
+      "step": 4080
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 16.108747482299805,
+      "learning_rate": 9.590590590590592e-06,
+      "loss": 1.0887,
+      "step": 4090
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 16.445880889892578,
+      "learning_rate": 9.58958958958959e-06,
+      "loss": 1.0269,
+      "step": 4100
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 13.307125091552734,
+      "learning_rate": 9.58858858858859e-06,
+      "loss": 1.0142,
+      "step": 4110
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 14.601872444152832,
+      "learning_rate": 9.587587587587588e-06,
+      "loss": 0.9902,
+      "step": 4120
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 18.373258590698242,
+      "learning_rate": 9.586586586586586e-06,
+      "loss": 1.0375,
+      "step": 4130
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 19.522417068481445,
+      "learning_rate": 9.585585585585587e-06,
+      "loss": 1.0825,
+      "step": 4140
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 18.246854782104492,
+      "learning_rate": 9.584584584584585e-06,
+      "loss": 0.9618,
+      "step": 4150
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 13.992356300354004,
+      "learning_rate": 9.583583583583584e-06,
+      "loss": 1.0104,
+      "step": 4160
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 12.092374801635742,
+      "learning_rate": 9.582582582582584e-06,
+      "loss": 1.0246,
+      "step": 4170
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 13.865076065063477,
+      "learning_rate": 9.581581581581583e-06,
+      "loss": 1.0414,
+      "step": 4180
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 15.241948127746582,
+      "learning_rate": 9.580580580580581e-06,
+      "loss": 0.9945,
+      "step": 4190
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 16.216447830200195,
+      "learning_rate": 9.57957957957958e-06,
+      "loss": 0.9745,
+      "step": 4200
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 15.302637100219727,
+      "learning_rate": 9.578578578578579e-06,
+      "loss": 0.987,
+      "step": 4210
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 18.61682891845703,
+      "learning_rate": 9.577577577577579e-06,
+      "loss": 0.9604,
+      "step": 4220
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 19.719160079956055,
+      "learning_rate": 9.576576576576578e-06,
+      "loss": 1.0032,
+      "step": 4230
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 18.480501174926758,
+      "learning_rate": 9.575575575575576e-06,
+      "loss": 1.0035,
+      "step": 4240
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 14.176997184753418,
+      "learning_rate": 9.574574574574575e-06,
+      "loss": 1.0165,
+      "step": 4250
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 20.608097076416016,
+      "learning_rate": 9.573573573573575e-06,
+      "loss": 0.9697,
+      "step": 4260
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 14.59773063659668,
+      "learning_rate": 9.572572572572574e-06,
+      "loss": 1.0698,
+      "step": 4270
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 12.92850112915039,
+      "learning_rate": 9.571571571571572e-06,
+      "loss": 1.0114,
+      "step": 4280
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 16.94991111755371,
+      "learning_rate": 9.57057057057057e-06,
+      "loss": 0.9624,
+      "step": 4290
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 14.561684608459473,
+      "learning_rate": 9.56956956956957e-06,
+      "loss": 1.0075,
+      "step": 4300
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 17.857587814331055,
+      "learning_rate": 9.56856856856857e-06,
+      "loss": 0.9379,
+      "step": 4310
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 22.47743034362793,
+      "learning_rate": 9.567567567567568e-06,
+      "loss": 1.0581,
+      "step": 4320
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.8379,
+      "eval_loss": 0.5319282412528992,
+      "eval_runtime": 12.9534,
+      "eval_samples_per_second": 772.0,
+      "eval_steps_per_second": 3.088,
+      "step": 4329
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 16.860095977783203,
+      "learning_rate": 9.566566566566567e-06,
+      "loss": 0.9651,
+      "step": 4330
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 14.949163436889648,
+      "learning_rate": 9.565565565565567e-06,
+      "loss": 0.933,
+      "step": 4340
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 18.923927307128906,
+      "learning_rate": 9.564564564564566e-06,
+      "loss": 1.0005,
+      "step": 4350
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 15.809991836547852,
+      "learning_rate": 9.563563563563564e-06,
+      "loss": 1.0425,
+      "step": 4360
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 18.985448837280273,
+      "learning_rate": 9.562562562562563e-06,
+      "loss": 1.0406,
+      "step": 4370
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 15.309414863586426,
+      "learning_rate": 9.561561561561562e-06,
+      "loss": 0.9674,
+      "step": 4380
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 18.867652893066406,
+      "learning_rate": 9.560560560560562e-06,
+      "loss": 1.0264,
+      "step": 4390
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 14.455560684204102,
+      "learning_rate": 9.55955955955956e-06,
+      "loss": 1.0368,
+      "step": 4400
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 19.578123092651367,
+      "learning_rate": 9.558558558558559e-06,
+      "loss": 0.9718,
+      "step": 4410
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 16.656957626342773,
+      "learning_rate": 9.55755755755756e-06,
+      "loss": 1.0459,
+      "step": 4420
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 28.463747024536133,
+      "learning_rate": 9.556556556556558e-06,
+      "loss": 1.006,
+      "step": 4430
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 20.07007598876953,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 0.9791,
+      "step": 4440
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 14.087189674377441,
+      "learning_rate": 9.554554554554555e-06,
+      "loss": 0.9722,
+      "step": 4450
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 18.301103591918945,
+      "learning_rate": 9.553553553553554e-06,
+      "loss": 1.0023,
+      "step": 4460
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 15.600085258483887,
+      "learning_rate": 9.552552552552552e-06,
+      "loss": 1.0245,
+      "step": 4470
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 11.453582763671875,
+      "learning_rate": 9.551551551551553e-06,
+      "loss": 0.9964,
+      "step": 4480
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 12.498774528503418,
+      "learning_rate": 9.550550550550551e-06,
+      "loss": 0.9411,
+      "step": 4490
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 14.276036262512207,
+      "learning_rate": 9.54954954954955e-06,
+      "loss": 0.9921,
+      "step": 4500
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 15.989923477172852,
+      "learning_rate": 9.54854854854855e-06,
+      "loss": 1.0208,
+      "step": 4510
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 12.61364459991455,
+      "learning_rate": 9.547547547547549e-06,
+      "loss": 0.9676,
+      "step": 4520
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 10.876855850219727,
+      "learning_rate": 9.546546546546547e-06,
+      "loss": 1.012,
+      "step": 4530
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 17.27198600769043,
+      "learning_rate": 9.545545545545546e-06,
+      "loss": 1.057,
+      "step": 4540
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 19.67862319946289,
+      "learning_rate": 9.544544544544544e-06,
+      "loss": 1.0207,
+      "step": 4550
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 15.878100395202637,
+      "learning_rate": 9.543543543543545e-06,
+      "loss": 1.0365,
+      "step": 4560
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 14.027839660644531,
+      "learning_rate": 9.542542542542543e-06,
+      "loss": 1.0017,
+      "step": 4570
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 16.443965911865234,
+      "learning_rate": 9.541541541541542e-06,
+      "loss": 0.9169,
+      "step": 4580
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 23.566343307495117,
+      "learning_rate": 9.540540540540542e-06,
+      "loss": 0.9557,
+      "step": 4590
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 14.526195526123047,
+      "learning_rate": 9.53953953953954e-06,
+      "loss": 0.9989,
+      "step": 4600
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 16.737215042114258,
+      "learning_rate": 9.53853853853854e-06,
+      "loss": 1.0153,
+      "step": 4610
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 14.807560920715332,
+      "learning_rate": 9.537537537537538e-06,
+      "loss": 0.8934,
+      "step": 4620
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 13.126487731933594,
+      "learning_rate": 9.536536536536537e-06,
+      "loss": 1.0443,
+      "step": 4630
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 20.58649253845215,
+      "learning_rate": 9.535535535535537e-06,
+      "loss": 1.0182,
+      "step": 4640
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 19.2255859375,
+      "learning_rate": 9.534534534534535e-06,
+      "loss": 0.93,
+      "step": 4650
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 13.996789932250977,
+      "learning_rate": 9.533533533533534e-06,
+      "loss": 0.9922,
+      "step": 4660
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.8388,
+      "eval_loss": 0.5215397477149963,
+      "eval_runtime": 12.6288,
+      "eval_samples_per_second": 791.843,
+      "eval_steps_per_second": 3.167,
+      "step": 4662
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 15.150822639465332,
+      "learning_rate": 9.532532532532534e-06,
+      "loss": 1.0928,
+      "step": 4670
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 17.099411010742188,
+      "learning_rate": 9.531531531531533e-06,
+      "loss": 0.9612,
+      "step": 4680
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 13.981841087341309,
+      "learning_rate": 9.530530530530532e-06,
+      "loss": 0.9182,
+      "step": 4690
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 20.836044311523438,
+      "learning_rate": 9.52952952952953e-06,
+      "loss": 0.953,
+      "step": 4700
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 18.01522445678711,
+      "learning_rate": 9.528528528528529e-06,
+      "loss": 0.9751,
+      "step": 4710
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 14.915099143981934,
+      "learning_rate": 9.527527527527527e-06,
+      "loss": 0.9972,
+      "step": 4720
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 19.267436981201172,
+      "learning_rate": 9.526526526526528e-06,
+      "loss": 1.0456,
+      "step": 4730
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 15.626789093017578,
+      "learning_rate": 9.525525525525526e-06,
+      "loss": 0.9834,
+      "step": 4740
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 17.810468673706055,
+      "learning_rate": 9.524524524524525e-06,
+      "loss": 0.9818,
+      "step": 4750
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 13.250941276550293,
+      "learning_rate": 9.523523523523525e-06,
+      "loss": 1.0015,
+      "step": 4760
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 19.812030792236328,
+      "learning_rate": 9.522522522522524e-06,
+      "loss": 0.9355,
+      "step": 4770
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 14.321195602416992,
+      "learning_rate": 9.521521521521522e-06,
+      "loss": 0.9828,
+      "step": 4780
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 18.426902770996094,
+      "learning_rate": 9.520520520520521e-06,
+      "loss": 0.9369,
+      "step": 4790
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 20.919023513793945,
+      "learning_rate": 9.51951951951952e-06,
+      "loss": 0.9663,
+      "step": 4800
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 15.742774963378906,
+      "learning_rate": 9.51851851851852e-06,
+      "loss": 1.0001,
+      "step": 4810
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 14.45275592803955,
+      "learning_rate": 9.517517517517518e-06,
+      "loss": 1.0518,
+      "step": 4820
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 21.9227294921875,
+      "learning_rate": 9.516516516516517e-06,
+      "loss": 1.0238,
+      "step": 4830
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 17.867456436157227,
+      "learning_rate": 9.515515515515517e-06,
+      "loss": 0.9458,
+      "step": 4840
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 19.42156219482422,
+      "learning_rate": 9.514514514514516e-06,
+      "loss": 0.9996,
+      "step": 4850
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 18.59282684326172,
+      "learning_rate": 9.513513513513514e-06,
+      "loss": 0.9686,
+      "step": 4860
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 15.362764358520508,
+      "learning_rate": 9.512512512512513e-06,
+      "loss": 0.9812,
+      "step": 4870
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 18.349945068359375,
+      "learning_rate": 9.511511511511512e-06,
+      "loss": 0.9779,
+      "step": 4880
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 13.918747901916504,
+      "learning_rate": 9.510510510510512e-06,
+      "loss": 0.9844,
+      "step": 4890
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 15.943765640258789,
+      "learning_rate": 9.50950950950951e-06,
+      "loss": 0.9398,
+      "step": 4900
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 11.988840103149414,
+      "learning_rate": 9.508508508508509e-06,
+      "loss": 0.9981,
+      "step": 4910
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 17.07158088684082,
+      "learning_rate": 9.507507507507508e-06,
+      "loss": 0.9832,
+      "step": 4920
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 15.438054084777832,
+      "learning_rate": 9.506506506506508e-06,
+      "loss": 0.9799,
+      "step": 4930
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 17.7360897064209,
+      "learning_rate": 9.505505505505507e-06,
+      "loss": 1.0055,
+      "step": 4940
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 14.747228622436523,
+      "learning_rate": 9.504504504504505e-06,
+      "loss": 1.0331,
+      "step": 4950
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 16.84065055847168,
+      "learning_rate": 9.503503503503504e-06,
+      "loss": 0.9135,
+      "step": 4960
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 18.45875358581543,
+      "learning_rate": 9.502502502502502e-06,
+      "loss": 1.0109,
+      "step": 4970
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 20.146686553955078,
+      "learning_rate": 9.501501501501503e-06,
+      "loss": 1.0193,
+      "step": 4980
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 15.921470642089844,
+      "learning_rate": 9.500500500500501e-06,
+      "loss": 0.9643,
+      "step": 4990
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.8399,
+      "eval_loss": 0.5144081115722656,
+      "eval_runtime": 12.647,
+      "eval_samples_per_second": 790.699,
+      "eval_steps_per_second": 3.163,
+      "step": 4995
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 14.36604118347168,
+      "learning_rate": 9.4994994994995e-06,
+      "loss": 0.8555,
+      "step": 5000
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 17.989551544189453,
+      "learning_rate": 9.4984984984985e-06,
+      "loss": 0.8889,
+      "step": 5010
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 15.000627517700195,
+      "learning_rate": 9.497497497497499e-06,
+      "loss": 0.9641,
+      "step": 5020
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 16.05202865600586,
+      "learning_rate": 9.496496496496497e-06,
+      "loss": 1.0126,
+      "step": 5030
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 16.95032501220703,
+      "learning_rate": 9.495495495495496e-06,
+      "loss": 0.9473,
+      "step": 5040
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 15.519562721252441,
+      "learning_rate": 9.494494494494494e-06,
+      "loss": 0.9128,
+      "step": 5050
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 20.99886131286621,
+      "learning_rate": 9.493493493493495e-06,
+      "loss": 0.9673,
+      "step": 5060
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 16.14808464050293,
+      "learning_rate": 9.492492492492493e-06,
+      "loss": 1.0206,
+      "step": 5070
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 14.916738510131836,
+      "learning_rate": 9.491491491491492e-06,
+      "loss": 0.9811,
+      "step": 5080
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 13.439791679382324,
+      "learning_rate": 9.490490490490492e-06,
+      "loss": 0.9779,
+      "step": 5090
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 15.340365409851074,
+      "learning_rate": 9.489489489489491e-06,
+      "loss": 0.9864,
+      "step": 5100
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 18.60448455810547,
+      "learning_rate": 9.48848848848849e-06,
+      "loss": 0.8913,
+      "step": 5110
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 11.516568183898926,
+      "learning_rate": 9.487487487487488e-06,
+      "loss": 0.9572,
+      "step": 5120
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 16.937572479248047,
+      "learning_rate": 9.486486486486487e-06,
+      "loss": 0.9264,
+      "step": 5130
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 13.213362693786621,
+      "learning_rate": 9.485485485485487e-06,
+      "loss": 0.8959,
+      "step": 5140
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 15.299351692199707,
+      "learning_rate": 9.484484484484486e-06,
+      "loss": 0.8851,
+      "step": 5150
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 14.688894271850586,
+      "learning_rate": 9.483483483483484e-06,
+      "loss": 0.9851,
+      "step": 5160
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 11.646239280700684,
+      "learning_rate": 9.482482482482483e-06,
+      "loss": 1.0083,
+      "step": 5170
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 18.091474533081055,
+      "learning_rate": 9.481481481481483e-06,
+      "loss": 0.8872,
+      "step": 5180
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 15.272651672363281,
+      "learning_rate": 9.480480480480482e-06,
+      "loss": 0.9658,
+      "step": 5190
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 21.172088623046875,
+      "learning_rate": 9.47947947947948e-06,
+      "loss": 0.9051,
+      "step": 5200
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 13.88657283782959,
+      "learning_rate": 9.478478478478479e-06,
+      "loss": 0.8966,
+      "step": 5210
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 14.709793090820312,
+      "learning_rate": 9.477477477477477e-06,
+      "loss": 0.8853,
+      "step": 5220
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 22.535512924194336,
+      "learning_rate": 9.476476476476478e-06,
+      "loss": 0.9747,
+      "step": 5230
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 20.59426498413086,
+      "learning_rate": 9.475475475475476e-06,
+      "loss": 0.9599,
+      "step": 5240
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 15.672638893127441,
+      "learning_rate": 9.474474474474475e-06,
+      "loss": 0.9429,
+      "step": 5250
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 21.314970016479492,
+      "learning_rate": 9.473473473473475e-06,
+      "loss": 0.9716,
+      "step": 5260
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 16.837343215942383,
+      "learning_rate": 9.472472472472474e-06,
+      "loss": 0.8394,
+      "step": 5270
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 15.07120132446289,
+      "learning_rate": 9.471471471471472e-06,
+      "loss": 0.9175,
+      "step": 5280
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 15.313618659973145,
+      "learning_rate": 9.470470470470471e-06,
+      "loss": 0.9212,
+      "step": 5290
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 21.591821670532227,
+      "learning_rate": 9.46946946946947e-06,
+      "loss": 0.9788,
+      "step": 5300
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 14.006278991699219,
+      "learning_rate": 9.46846846846847e-06,
+      "loss": 0.9606,
+      "step": 5310
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 13.330794334411621,
+      "learning_rate": 9.467467467467468e-06,
+      "loss": 0.9687,
+      "step": 5320
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.8413,
+      "eval_loss": 0.5102618932723999,
+      "eval_runtime": 12.8842,
+      "eval_samples_per_second": 776.145,
+      "eval_steps_per_second": 3.105,
+      "step": 5328
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 18.027389526367188,
+      "learning_rate": 9.466466466466467e-06,
+      "loss": 1.1231,
+      "step": 5330
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 16.487071990966797,
+      "learning_rate": 9.465465465465467e-06,
+      "loss": 0.9447,
+      "step": 5340
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 14.945123672485352,
+      "learning_rate": 9.464464464464466e-06,
+      "loss": 0.9224,
+      "step": 5350
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 23.017677307128906,
+      "learning_rate": 9.463463463463464e-06,
+      "loss": 0.9404,
+      "step": 5360
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 16.184133529663086,
+      "learning_rate": 9.462462462462463e-06,
+      "loss": 0.9412,
+      "step": 5370
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 15.728362083435059,
+      "learning_rate": 9.461461461461462e-06,
+      "loss": 0.9546,
+      "step": 5380
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 16.84214973449707,
+      "learning_rate": 9.46046046046046e-06,
+      "loss": 0.9157,
+      "step": 5390
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 15.392135620117188,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.9431,
+      "step": 5400
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 17.416915893554688,
+      "learning_rate": 9.458458458458459e-06,
+      "loss": 0.9596,
+      "step": 5410
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 14.695914268493652,
+      "learning_rate": 9.457457457457458e-06,
+      "loss": 0.9424,
+      "step": 5420
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 14.552547454833984,
+      "learning_rate": 9.456456456456458e-06,
+      "loss": 0.9052,
+      "step": 5430
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 21.200580596923828,
+      "learning_rate": 9.455455455455457e-06,
+      "loss": 0.9945,
+      "step": 5440
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 20.810646057128906,
+      "learning_rate": 9.454454454454455e-06,
+      "loss": 0.9052,
+      "step": 5450
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 21.23904037475586,
+      "learning_rate": 9.453453453453454e-06,
+      "loss": 0.98,
+      "step": 5460
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 14.602259635925293,
+      "learning_rate": 9.452452452452452e-06,
+      "loss": 0.9851,
+      "step": 5470
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 16.075517654418945,
+      "learning_rate": 9.451451451451453e-06,
+      "loss": 1.0038,
+      "step": 5480
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 13.073135375976562,
+      "learning_rate": 9.450450450450451e-06,
+      "loss": 0.9587,
+      "step": 5490
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 16.11018180847168,
+      "learning_rate": 9.44944944944945e-06,
+      "loss": 0.9649,
+      "step": 5500
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 17.22406578063965,
+      "learning_rate": 9.44844844844845e-06,
+      "loss": 0.8726,
+      "step": 5510
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 12.76385498046875,
+      "learning_rate": 9.447447447447449e-06,
+      "loss": 0.893,
+      "step": 5520
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 16.00253677368164,
+      "learning_rate": 9.446446446446447e-06,
+      "loss": 0.9512,
+      "step": 5530
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 15.32001781463623,
+      "learning_rate": 9.445445445445446e-06,
+      "loss": 0.9346,
+      "step": 5540
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 17.619962692260742,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.9255,
+      "step": 5550
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 18.094051361083984,
+      "learning_rate": 9.443443443443445e-06,
+      "loss": 0.942,
+      "step": 5560
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 14.032143592834473,
+      "learning_rate": 9.442442442442443e-06,
+      "loss": 0.8866,
+      "step": 5570
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 13.485750198364258,
+      "learning_rate": 9.441441441441442e-06,
+      "loss": 0.9054,
+      "step": 5580
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 18.393156051635742,
+      "learning_rate": 9.440440440440442e-06,
+      "loss": 0.9418,
+      "step": 5590
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 21.527530670166016,
+      "learning_rate": 9.439439439439441e-06,
+      "loss": 0.9645,
+      "step": 5600
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 16.544702529907227,
+      "learning_rate": 9.43843843843844e-06,
+      "loss": 0.9386,
+      "step": 5610
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 15.755117416381836,
+      "learning_rate": 9.437437437437438e-06,
+      "loss": 0.9343,
+      "step": 5620
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 19.206357955932617,
+      "learning_rate": 9.436436436436437e-06,
+      "loss": 0.9381,
+      "step": 5630
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 16.807981491088867,
+      "learning_rate": 9.435435435435435e-06,
+      "loss": 0.8847,
+      "step": 5640
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 14.00562572479248,
+      "learning_rate": 9.434434434434436e-06,
+      "loss": 1.0009,
+      "step": 5650
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 25.151762008666992,
+      "learning_rate": 9.433433433433434e-06,
+      "loss": 0.9464,
+      "step": 5660
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.8422,
+      "eval_loss": 0.5020662546157837,
+      "eval_runtime": 12.5537,
+      "eval_samples_per_second": 796.58,
+      "eval_steps_per_second": 3.186,
+      "step": 5661
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 15.472583770751953,
+      "learning_rate": 9.432432432432433e-06,
+      "loss": 1.0577,
+      "step": 5670
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 16.592697143554688,
+      "learning_rate": 9.431431431431433e-06,
+      "loss": 0.9467,
+      "step": 5680
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 22.673389434814453,
+      "learning_rate": 9.430430430430432e-06,
+      "loss": 0.9585,
+      "step": 5690
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 17.902254104614258,
+      "learning_rate": 9.42942942942943e-06,
+      "loss": 0.8856,
+      "step": 5700
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 20.870149612426758,
+      "learning_rate": 9.428428428428429e-06,
+      "loss": 0.9163,
+      "step": 5710
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 17.17050552368164,
+      "learning_rate": 9.427427427427427e-06,
+      "loss": 0.9319,
+      "step": 5720
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 15.78785514831543,
+      "learning_rate": 9.426426426426428e-06,
+      "loss": 0.8843,
+      "step": 5730
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 15.386320114135742,
+      "learning_rate": 9.425425425425426e-06,
+      "loss": 0.8933,
+      "step": 5740
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 19.489320755004883,
+      "learning_rate": 9.424424424424425e-06,
+      "loss": 0.8776,
+      "step": 5750
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 34.896873474121094,
+      "learning_rate": 9.423423423423425e-06,
+      "loss": 0.9789,
+      "step": 5760
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 11.27894401550293,
+      "learning_rate": 9.422422422422424e-06,
+      "loss": 0.9901,
+      "step": 5770
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 20.736225128173828,
+      "learning_rate": 9.421421421421422e-06,
+      "loss": 0.9602,
+      "step": 5780
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 14.630104064941406,
+      "learning_rate": 9.420420420420421e-06,
+      "loss": 0.8607,
+      "step": 5790
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 16.037254333496094,
+      "learning_rate": 9.41941941941942e-06,
+      "loss": 0.9457,
+      "step": 5800
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 18.13541603088379,
+      "learning_rate": 9.41841841841842e-06,
+      "loss": 0.9498,
+      "step": 5810
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 11.522095680236816,
+      "learning_rate": 9.417417417417418e-06,
+      "loss": 0.8653,
+      "step": 5820
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 16.0896053314209,
+      "learning_rate": 9.416416416416417e-06,
+      "loss": 0.9219,
+      "step": 5830
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 13.031556129455566,
+      "learning_rate": 9.415415415415416e-06,
+      "loss": 0.8919,
+      "step": 5840
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 13.607986450195312,
+      "learning_rate": 9.414414414414416e-06,
+      "loss": 0.9314,
+      "step": 5850
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 21.58922004699707,
+      "learning_rate": 9.413413413413413e-06,
+      "loss": 0.887,
+      "step": 5860
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 17.312057495117188,
+      "learning_rate": 9.412412412412413e-06,
+      "loss": 0.9299,
+      "step": 5870
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 17.167619705200195,
+      "learning_rate": 9.411411411411412e-06,
+      "loss": 0.9754,
+      "step": 5880
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 17.25926399230957,
+      "learning_rate": 9.41041041041041e-06,
+      "loss": 0.8431,
+      "step": 5890
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 24.359844207763672,
+      "learning_rate": 9.40940940940941e-06,
+      "loss": 0.9661,
+      "step": 5900
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 17.767169952392578,
+      "learning_rate": 9.40840840840841e-06,
+      "loss": 0.9478,
+      "step": 5910
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 18.630041122436523,
+      "learning_rate": 9.407407407407408e-06,
+      "loss": 0.8989,
+      "step": 5920
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 13.352246284484863,
+      "learning_rate": 9.406406406406408e-06,
+      "loss": 0.964,
+      "step": 5930
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 14.739523887634277,
+      "learning_rate": 9.405405405405407e-06,
+      "loss": 0.9195,
+      "step": 5940
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 18.194190979003906,
+      "learning_rate": 9.404404404404405e-06,
+      "loss": 0.9098,
+      "step": 5950
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 20.455427169799805,
+      "learning_rate": 9.403403403403404e-06,
+      "loss": 0.8647,
+      "step": 5960
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 17.527029037475586,
+      "learning_rate": 9.402402402402402e-06,
+      "loss": 0.967,
+      "step": 5970
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 16.669889450073242,
+      "learning_rate": 9.401401401401403e-06,
+      "loss": 0.9095,
+      "step": 5980
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 16.57728385925293,
+      "learning_rate": 9.400400400400401e-06,
+      "loss": 0.8651,
+      "step": 5990
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.8483,
+      "eval_loss": 0.4867164194583893,
+      "eval_runtime": 12.5556,
+      "eval_samples_per_second": 796.458,
+      "eval_steps_per_second": 3.186,
+      "step": 5994
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 16.830820083618164,
+      "learning_rate": 9.3993993993994e-06,
+      "loss": 0.8231,
+      "step": 6000
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 13.973091125488281,
+      "learning_rate": 9.3983983983984e-06,
+      "loss": 0.8579,
+      "step": 6010
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 26.990978240966797,
+      "learning_rate": 9.397397397397399e-06,
+      "loss": 0.9031,
+      "step": 6020
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 19.440040588378906,
+      "learning_rate": 9.396396396396397e-06,
+      "loss": 0.8558,
+      "step": 6030
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 16.811927795410156,
+      "learning_rate": 9.395395395395396e-06,
+      "loss": 0.8856,
+      "step": 6040
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 15.919425964355469,
+      "learning_rate": 9.394394394394395e-06,
+      "loss": 0.8952,
+      "step": 6050
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 14.041596412658691,
+      "learning_rate": 9.393393393393393e-06,
+      "loss": 0.8978,
+      "step": 6060
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 15.746431350708008,
+      "learning_rate": 9.392392392392394e-06,
+      "loss": 0.9087,
+      "step": 6070
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 11.961185455322266,
+      "learning_rate": 9.391391391391392e-06,
+      "loss": 0.9387,
+      "step": 6080
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 18.50293731689453,
+      "learning_rate": 9.39039039039039e-06,
+      "loss": 0.9178,
+      "step": 6090
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 16.186243057250977,
+      "learning_rate": 9.389389389389391e-06,
+      "loss": 0.9176,
+      "step": 6100
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 16.879858016967773,
+      "learning_rate": 9.388388388388388e-06,
+      "loss": 0.8795,
+      "step": 6110
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 14.703543663024902,
+      "learning_rate": 9.387387387387388e-06,
+      "loss": 0.8627,
+      "step": 6120
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 14.977249145507812,
+      "learning_rate": 9.386386386386387e-06,
+      "loss": 0.9203,
+      "step": 6130
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 23.646411895751953,
+      "learning_rate": 9.385385385385385e-06,
+      "loss": 0.8878,
+      "step": 6140
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 20.65953254699707,
+      "learning_rate": 9.384384384384386e-06,
+      "loss": 0.8367,
+      "step": 6150
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 15.61327838897705,
+      "learning_rate": 9.383383383383384e-06,
+      "loss": 0.8978,
+      "step": 6160
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 15.82312297821045,
+      "learning_rate": 9.382382382382383e-06,
+      "loss": 0.9158,
+      "step": 6170
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 15.94821548461914,
+      "learning_rate": 9.381381381381383e-06,
+      "loss": 0.9168,
+      "step": 6180
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 14.084745407104492,
+      "learning_rate": 9.380380380380382e-06,
+      "loss": 0.914,
+      "step": 6190
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 14.570650100708008,
+      "learning_rate": 9.37937937937938e-06,
+      "loss": 0.8604,
+      "step": 6200
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 14.09744930267334,
+      "learning_rate": 9.378378378378379e-06,
+      "loss": 0.8786,
+      "step": 6210
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 18.788267135620117,
+      "learning_rate": 9.377377377377378e-06,
+      "loss": 0.9293,
+      "step": 6220
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 15.376623153686523,
+      "learning_rate": 9.376376376376378e-06,
+      "loss": 0.8625,
+      "step": 6230
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 17.357419967651367,
+      "learning_rate": 9.375375375375376e-06,
+      "loss": 0.9037,
+      "step": 6240
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 13.640501022338867,
+      "learning_rate": 9.374374374374375e-06,
+      "loss": 0.9497,
+      "step": 6250
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 15.700154304504395,
+      "learning_rate": 9.373373373373375e-06,
+      "loss": 0.9401,
+      "step": 6260
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 17.42572784423828,
+      "learning_rate": 9.372372372372374e-06,
+      "loss": 0.9247,
+      "step": 6270
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 17.637113571166992,
+      "learning_rate": 9.371371371371372e-06,
+      "loss": 0.917,
+      "step": 6280
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 13.902384757995605,
+      "learning_rate": 9.370370370370371e-06,
+      "loss": 0.8463,
+      "step": 6290
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 16.326753616333008,
+      "learning_rate": 9.36936936936937e-06,
+      "loss": 0.8557,
+      "step": 6300
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 14.994547843933105,
+      "learning_rate": 9.368368368368368e-06,
+      "loss": 0.9127,
+      "step": 6310
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 23.689701080322266,
+      "learning_rate": 9.367367367367369e-06,
+      "loss": 0.8122,
+      "step": 6320
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.8457,
+      "eval_loss": 0.48648661375045776,
+      "eval_runtime": 12.9306,
+      "eval_samples_per_second": 773.358,
+      "eval_steps_per_second": 3.093,
+      "step": 6327
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 14.423419952392578,
+      "learning_rate": 9.366366366366367e-06,
+      "loss": 0.9854,
+      "step": 6330
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 18.28612518310547,
+      "learning_rate": 9.365365365365366e-06,
+      "loss": 0.9195,
+      "step": 6340
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 14.99592399597168,
+      "learning_rate": 9.364364364364366e-06,
+      "loss": 0.8935,
+      "step": 6350
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 16.575809478759766,
+      "learning_rate": 9.363363363363363e-06,
+      "loss": 0.9202,
+      "step": 6360
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 21.606903076171875,
+      "learning_rate": 9.362362362362363e-06,
+      "loss": 0.9039,
+      "step": 6370
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 14.512263298034668,
+      "learning_rate": 9.361361361361362e-06,
+      "loss": 0.8536,
+      "step": 6380
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 19.58755874633789,
+      "learning_rate": 9.36036036036036e-06,
+      "loss": 0.839,
+      "step": 6390
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 22.35240364074707,
+      "learning_rate": 9.35935935935936e-06,
+      "loss": 0.8474,
+      "step": 6400
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 12.306565284729004,
+      "learning_rate": 9.35835835835836e-06,
+      "loss": 0.927,
+      "step": 6410
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 19.176259994506836,
+      "learning_rate": 9.357357357357358e-06,
+      "loss": 0.9118,
+      "step": 6420
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 16.825773239135742,
+      "learning_rate": 9.356356356356358e-06,
+      "loss": 0.9009,
+      "step": 6430
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 23.165769577026367,
+      "learning_rate": 9.355355355355357e-06,
+      "loss": 0.8732,
+      "step": 6440
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 13.343769073486328,
+      "learning_rate": 9.354354354354355e-06,
+      "loss": 0.9431,
+      "step": 6450
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 13.997703552246094,
+      "learning_rate": 9.353353353353354e-06,
+      "loss": 0.8847,
+      "step": 6460
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 16.435840606689453,
+      "learning_rate": 9.352352352352353e-06,
+      "loss": 0.8549,
+      "step": 6470
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 15.106839179992676,
+      "learning_rate": 9.351351351351353e-06,
+      "loss": 0.8082,
+      "step": 6480
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 18.501834869384766,
+      "learning_rate": 9.350350350350351e-06,
+      "loss": 0.9178,
+      "step": 6490
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 16.704853057861328,
+      "learning_rate": 9.34934934934935e-06,
+      "loss": 0.8601,
+      "step": 6500
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 13.834967613220215,
+      "learning_rate": 9.34834834834835e-06,
+      "loss": 0.9497,
+      "step": 6510
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 14.41886043548584,
+      "learning_rate": 9.347347347347349e-06,
+      "loss": 0.8517,
+      "step": 6520
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 18.210779190063477,
+      "learning_rate": 9.346346346346346e-06,
+      "loss": 0.8575,
+      "step": 6530
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 13.764591217041016,
+      "learning_rate": 9.345345345345346e-06,
+      "loss": 0.9236,
+      "step": 6540
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 19.188827514648438,
+      "learning_rate": 9.344344344344345e-06,
+      "loss": 0.8965,
+      "step": 6550
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 16.712377548217773,
+      "learning_rate": 9.343343343343343e-06,
+      "loss": 0.8958,
+      "step": 6560
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 16.866111755371094,
+      "learning_rate": 9.342342342342344e-06,
+      "loss": 0.8984,
+      "step": 6570
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 16.404024124145508,
+      "learning_rate": 9.341341341341342e-06,
+      "loss": 0.8949,
+      "step": 6580
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 14.78878402709961,
+      "learning_rate": 9.34034034034034e-06,
+      "loss": 0.852,
+      "step": 6590
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 19.638397216796875,
+      "learning_rate": 9.339339339339341e-06,
+      "loss": 0.9375,
+      "step": 6600
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 15.815281867980957,
+      "learning_rate": 9.338338338338338e-06,
+      "loss": 0.922,
+      "step": 6610
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 12.827592849731445,
+      "learning_rate": 9.337337337337338e-06,
+      "loss": 0.8804,
+      "step": 6620
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 16.541942596435547,
+      "learning_rate": 9.336336336336337e-06,
+      "loss": 0.9358,
+      "step": 6630
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 17.47844123840332,
+      "learning_rate": 9.335335335335335e-06,
+      "loss": 0.9289,
+      "step": 6640
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 15.426770210266113,
+      "learning_rate": 9.334334334334336e-06,
+      "loss": 0.8539,
+      "step": 6650
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 25.500600814819336,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.7918,
+      "step": 6660
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.8486,
+      "eval_loss": 0.4876756966114044,
+      "eval_runtime": 12.6548,
+      "eval_samples_per_second": 790.216,
+      "eval_steps_per_second": 3.161,
+      "step": 6660
+    },
+    {
+      "epoch": 20.03,
+      "grad_norm": 20.535430908203125,
+      "learning_rate": 9.332332332332333e-06,
+      "loss": 0.8509,
+      "step": 6670
+    },
+    {
+      "epoch": 20.06,
+      "grad_norm": 15.317164421081543,
+      "learning_rate": 9.331331331331333e-06,
+      "loss": 0.8385,
+      "step": 6680
+    },
+    {
+      "epoch": 20.09,
+      "grad_norm": 21.33066749572754,
+      "learning_rate": 9.330330330330332e-06,
+      "loss": 0.8901,
+      "step": 6690
+    },
+    {
+      "epoch": 20.12,
+      "grad_norm": 13.255439758300781,
+      "learning_rate": 9.32932932932933e-06,
+      "loss": 0.8398,
+      "step": 6700
+    },
+    {
+      "epoch": 20.15,
+      "grad_norm": 17.686439514160156,
+      "learning_rate": 9.328328328328329e-06,
+      "loss": 0.716,
+      "step": 6710
+    },
+    {
+      "epoch": 20.18,
+      "grad_norm": 17.935535430908203,
+      "learning_rate": 9.327327327327328e-06,
+      "loss": 0.8567,
+      "step": 6720
+    },
+    {
+      "epoch": 20.21,
+      "grad_norm": 12.795973777770996,
+      "learning_rate": 9.326326326326328e-06,
+      "loss": 0.876,
+      "step": 6730
+    },
+    {
+      "epoch": 20.24,
+      "grad_norm": 12.963747024536133,
+      "learning_rate": 9.325325325325326e-06,
+      "loss": 0.8559,
+      "step": 6740
+    },
+    {
+      "epoch": 20.27,
+      "grad_norm": 20.37497329711914,
+      "learning_rate": 9.324324324324325e-06,
+      "loss": 0.9035,
+      "step": 6750
+    },
+    {
+      "epoch": 20.3,
+      "grad_norm": 15.318814277648926,
+      "learning_rate": 9.323323323323324e-06,
+      "loss": 0.8254,
+      "step": 6760
+    },
+    {
+      "epoch": 20.33,
+      "grad_norm": 14.240460395812988,
+      "learning_rate": 9.322322322322324e-06,
+      "loss": 0.8645,
+      "step": 6770
+    },
+    {
+      "epoch": 20.36,
+      "grad_norm": 17.571277618408203,
+      "learning_rate": 9.321321321321321e-06,
+      "loss": 0.9952,
+      "step": 6780
+    },
+    {
+      "epoch": 20.39,
+      "grad_norm": 22.65481948852539,
+      "learning_rate": 9.320320320320321e-06,
+      "loss": 0.8574,
+      "step": 6790
+    },
+    {
+      "epoch": 20.42,
+      "grad_norm": 20.18941879272461,
+      "learning_rate": 9.31931931931932e-06,
+      "loss": 0.9161,
+      "step": 6800
+    },
+    {
+      "epoch": 20.45,
+      "grad_norm": 15.671890258789062,
+      "learning_rate": 9.318318318318318e-06,
+      "loss": 0.8537,
+      "step": 6810
+    },
+    {
+      "epoch": 20.48,
+      "grad_norm": 15.43073844909668,
+      "learning_rate": 9.317317317317319e-06,
+      "loss": 0.9022,
+      "step": 6820
+    },
+    {
+      "epoch": 20.51,
+      "grad_norm": 12.564041137695312,
+      "learning_rate": 9.316316316316317e-06,
+      "loss": 0.858,
+      "step": 6830
+    },
+    {
+      "epoch": 20.54,
+      "grad_norm": 14.09218978881836,
+      "learning_rate": 9.315315315315316e-06,
+      "loss": 0.8307,
+      "step": 6840
+    },
+    {
+      "epoch": 20.57,
+      "grad_norm": 17.242481231689453,
+      "learning_rate": 9.314314314314316e-06,
+      "loss": 0.8369,
+      "step": 6850
+    },
+    {
+      "epoch": 20.6,
+      "grad_norm": 14.80042552947998,
+      "learning_rate": 9.313313313313313e-06,
+      "loss": 0.9467,
+      "step": 6860
+    },
+    {
+      "epoch": 20.63,
+      "grad_norm": 14.18408203125,
+      "learning_rate": 9.312312312312313e-06,
+      "loss": 0.8499,
+      "step": 6870
+    },
+    {
+      "epoch": 20.66,
+      "grad_norm": 14.217787742614746,
+      "learning_rate": 9.311311311311312e-06,
+      "loss": 0.8993,
+      "step": 6880
+    },
+    {
+      "epoch": 20.69,
+      "grad_norm": 15.242293357849121,
+      "learning_rate": 9.31031031031031e-06,
+      "loss": 0.8853,
+      "step": 6890
+    },
+    {
+      "epoch": 20.72,
+      "grad_norm": 12.323200225830078,
+      "learning_rate": 9.30930930930931e-06,
+      "loss": 0.8234,
+      "step": 6900
+    },
+    {
+      "epoch": 20.75,
+      "grad_norm": 17.57102394104004,
+      "learning_rate": 9.30830830830831e-06,
+      "loss": 0.8901,
+      "step": 6910
+    },
+    {
+      "epoch": 20.78,
+      "grad_norm": 10.80214786529541,
+      "learning_rate": 9.307307307307308e-06,
+      "loss": 0.8077,
+      "step": 6920
+    },
+    {
+      "epoch": 20.81,
+      "grad_norm": 11.871294021606445,
+      "learning_rate": 9.306306306306308e-06,
+      "loss": 0.8732,
+      "step": 6930
+    },
+    {
+      "epoch": 20.84,
+      "grad_norm": 14.237271308898926,
+      "learning_rate": 9.305305305305305e-06,
+      "loss": 0.8879,
+      "step": 6940
+    },
+    {
+      "epoch": 20.87,
+      "grad_norm": 16.12568473815918,
+      "learning_rate": 9.304304304304305e-06,
+      "loss": 0.9051,
+      "step": 6950
+    },
+    {
+      "epoch": 20.9,
+      "grad_norm": 16.860570907592773,
+      "learning_rate": 9.303303303303304e-06,
+      "loss": 0.9073,
+      "step": 6960
+    },
+    {
+      "epoch": 20.93,
+      "grad_norm": 13.383695602416992,
+      "learning_rate": 9.302302302302303e-06,
+      "loss": 0.8684,
+      "step": 6970
+    },
+    {
+      "epoch": 20.96,
+      "grad_norm": 16.58403205871582,
+      "learning_rate": 9.301301301301301e-06,
+      "loss": 0.8682,
+      "step": 6980
+    },
+    {
+      "epoch": 20.99,
+      "grad_norm": 14.263192176818848,
+      "learning_rate": 9.300300300300302e-06,
+      "loss": 0.8994,
+      "step": 6990
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.8502,
+      "eval_loss": 0.4836479425430298,
+      "eval_runtime": 12.9867,
+      "eval_samples_per_second": 770.017,
+      "eval_steps_per_second": 3.08,
+      "step": 6993
+    },
+    {
+      "epoch": 21.02,
+      "grad_norm": 14.633737564086914,
+      "learning_rate": 9.2992992992993e-06,
+      "loss": 0.9701,
+      "step": 7000
+    },
+    {
+      "epoch": 21.05,
+      "grad_norm": 16.935035705566406,
+      "learning_rate": 9.298298298298299e-06,
+      "loss": 0.8863,
+      "step": 7010
+    },
+    {
+      "epoch": 21.08,
+      "grad_norm": 14.481687545776367,
+      "learning_rate": 9.297297297297299e-06,
+      "loss": 0.84,
+      "step": 7020
+    },
+    {
+      "epoch": 21.11,
+      "grad_norm": 11.950042724609375,
+      "learning_rate": 9.296296296296296e-06,
+      "loss": 0.794,
+      "step": 7030
+    },
+    {
+      "epoch": 21.14,
+      "grad_norm": 18.95380210876465,
+      "learning_rate": 9.295295295295296e-06,
+      "loss": 0.8477,
+      "step": 7040
+    },
+    {
+      "epoch": 21.17,
+      "grad_norm": 13.028655052185059,
+      "learning_rate": 9.294294294294295e-06,
+      "loss": 0.8056,
+      "step": 7050
+    },
+    {
+      "epoch": 21.2,
+      "grad_norm": 16.111392974853516,
+      "learning_rate": 9.293293293293293e-06,
+      "loss": 0.8293,
+      "step": 7060
+    },
+    {
+      "epoch": 21.23,
+      "grad_norm": 28.997385025024414,
+      "learning_rate": 9.292292292292294e-06,
+      "loss": 0.8456,
+      "step": 7070
+    },
+    {
+      "epoch": 21.26,
+      "grad_norm": 16.135433197021484,
+      "learning_rate": 9.291291291291292e-06,
+      "loss": 0.8633,
+      "step": 7080
+    },
+    {
+      "epoch": 21.29,
+      "grad_norm": 18.188039779663086,
+      "learning_rate": 9.29029029029029e-06,
+      "loss": 0.9532,
+      "step": 7090
+    },
+    {
+      "epoch": 21.32,
+      "grad_norm": 18.010501861572266,
+      "learning_rate": 9.289289289289291e-06,
+      "loss": 0.7915,
+      "step": 7100
+    },
+    {
+      "epoch": 21.35,
+      "grad_norm": 17.852554321289062,
+      "learning_rate": 9.288288288288288e-06,
+      "loss": 0.8928,
+      "step": 7110
+    },
+    {
+      "epoch": 21.38,
+      "grad_norm": 15.39466667175293,
+      "learning_rate": 9.287287287287288e-06,
+      "loss": 0.7896,
+      "step": 7120
+    },
+    {
+      "epoch": 21.41,
+      "grad_norm": 16.305221557617188,
+      "learning_rate": 9.286286286286287e-06,
+      "loss": 0.9138,
+      "step": 7130
+    },
+    {
+      "epoch": 21.44,
+      "grad_norm": 14.550299644470215,
+      "learning_rate": 9.285285285285286e-06,
+      "loss": 0.916,
+      "step": 7140
+    },
+    {
+      "epoch": 21.47,
+      "grad_norm": 18.119386672973633,
+      "learning_rate": 9.284284284284286e-06,
+      "loss": 0.8451,
+      "step": 7150
+    },
+    {
+      "epoch": 21.5,
+      "grad_norm": 21.804479598999023,
+      "learning_rate": 9.283283283283284e-06,
+      "loss": 0.887,
+      "step": 7160
+    },
+    {
+      "epoch": 21.53,
+      "grad_norm": 15.080596923828125,
+      "learning_rate": 9.282282282282283e-06,
+      "loss": 0.9324,
+      "step": 7170
+    },
+    {
+      "epoch": 21.56,
+      "grad_norm": 15.653203010559082,
+      "learning_rate": 9.281281281281283e-06,
+      "loss": 0.8725,
+      "step": 7180
+    },
+    {
+      "epoch": 21.59,
+      "grad_norm": 13.02376937866211,
+      "learning_rate": 9.28028028028028e-06,
+      "loss": 0.8557,
+      "step": 7190
+    },
+    {
+      "epoch": 21.62,
+      "grad_norm": 16.726030349731445,
+      "learning_rate": 9.27927927927928e-06,
+      "loss": 0.804,
+      "step": 7200
+    },
+    {
+      "epoch": 21.65,
+      "grad_norm": 24.801246643066406,
+      "learning_rate": 9.278278278278279e-06,
+      "loss": 0.895,
+      "step": 7210
+    },
+    {
+      "epoch": 21.68,
+      "grad_norm": 18.446496963500977,
+      "learning_rate": 9.277277277277278e-06,
+      "loss": 0.9239,
+      "step": 7220
+    },
+    {
+      "epoch": 21.71,
+      "grad_norm": 19.611867904663086,
+      "learning_rate": 9.276276276276276e-06,
+      "loss": 0.7734,
+      "step": 7230
+    },
+    {
+      "epoch": 21.74,
+      "grad_norm": 17.296485900878906,
+      "learning_rate": 9.275275275275277e-06,
+      "loss": 0.8745,
+      "step": 7240
+    },
+    {
+      "epoch": 21.77,
+      "grad_norm": 22.829599380493164,
+      "learning_rate": 9.274274274274275e-06,
+      "loss": 0.8779,
+      "step": 7250
+    },
+    {
+      "epoch": 21.8,
+      "grad_norm": 17.0224552154541,
+      "learning_rate": 9.273273273273274e-06,
+      "loss": 0.8653,
+      "step": 7260
+    },
+    {
+      "epoch": 21.83,
+      "grad_norm": 19.823801040649414,
+      "learning_rate": 9.272272272272274e-06,
+      "loss": 0.8803,
+      "step": 7270
+    },
+    {
+      "epoch": 21.86,
+      "grad_norm": 13.497579574584961,
+      "learning_rate": 9.271271271271271e-06,
+      "loss": 0.8718,
+      "step": 7280
+    },
+    {
+      "epoch": 21.89,
+      "grad_norm": 13.915096282958984,
+      "learning_rate": 9.270270270270271e-06,
+      "loss": 0.8566,
+      "step": 7290
+    },
+    {
+      "epoch": 21.92,
+      "grad_norm": 19.492271423339844,
+      "learning_rate": 9.26926926926927e-06,
+      "loss": 0.8285,
+      "step": 7300
+    },
+    {
+      "epoch": 21.95,
+      "grad_norm": 11.919459342956543,
+      "learning_rate": 9.268268268268268e-06,
+      "loss": 0.859,
+      "step": 7310
+    },
+    {
+      "epoch": 21.98,
+      "grad_norm": 12.806656837463379,
+      "learning_rate": 9.267267267267269e-06,
+      "loss": 0.8661,
+      "step": 7320
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.8538,
+      "eval_loss": 0.47356659173965454,
+      "eval_runtime": 12.8169,
+      "eval_samples_per_second": 780.217,
+      "eval_steps_per_second": 3.121,
+      "step": 7326
+    },
+    {
+      "epoch": 22.01,
+      "grad_norm": 13.644493103027344,
+      "learning_rate": 9.266266266266267e-06,
+      "loss": 0.9675,
+      "step": 7330
+    },
+    {
+      "epoch": 22.04,
+      "grad_norm": 18.335735321044922,
+      "learning_rate": 9.265265265265266e-06,
+      "loss": 0.8491,
+      "step": 7340
+    },
+    {
+      "epoch": 22.07,
+      "grad_norm": 15.356688499450684,
+      "learning_rate": 9.264264264264266e-06,
+      "loss": 0.85,
+      "step": 7350
+    },
+    {
+      "epoch": 22.1,
+      "grad_norm": 16.97271156311035,
+      "learning_rate": 9.263263263263263e-06,
+      "loss": 0.8038,
+      "step": 7360
+    },
+    {
+      "epoch": 22.13,
+      "grad_norm": 15.182245254516602,
+      "learning_rate": 9.262262262262263e-06,
+      "loss": 0.8134,
+      "step": 7370
+    },
+    {
+      "epoch": 22.16,
+      "grad_norm": 15.863039016723633,
+      "learning_rate": 9.261261261261262e-06,
+      "loss": 0.7958,
+      "step": 7380
+    },
+    {
+      "epoch": 22.19,
+      "grad_norm": 11.823266983032227,
+      "learning_rate": 9.26026026026026e-06,
+      "loss": 0.8699,
+      "step": 7390
+    },
+    {
+      "epoch": 22.22,
+      "grad_norm": 14.815451622009277,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 0.8234,
+      "step": 7400
+    },
+    {
+      "epoch": 22.25,
+      "grad_norm": 20.885053634643555,
+      "learning_rate": 9.25825825825826e-06,
+      "loss": 0.8443,
+      "step": 7410
+    },
+    {
+      "epoch": 22.28,
+      "grad_norm": 14.285574913024902,
+      "learning_rate": 9.257257257257258e-06,
+      "loss": 0.845,
+      "step": 7420
+    },
+    {
+      "epoch": 22.31,
+      "grad_norm": 17.085548400878906,
+      "learning_rate": 9.256256256256257e-06,
+      "loss": 0.8106,
+      "step": 7430
+    },
+    {
+      "epoch": 22.34,
+      "grad_norm": 27.535512924194336,
+      "learning_rate": 9.255255255255255e-06,
+      "loss": 0.8112,
+      "step": 7440
+    },
+    {
+      "epoch": 22.37,
+      "grad_norm": 15.75393295288086,
+      "learning_rate": 9.254254254254254e-06,
+      "loss": 0.8391,
+      "step": 7450
+    },
+    {
+      "epoch": 22.4,
+      "grad_norm": 22.823278427124023,
+      "learning_rate": 9.253253253253254e-06,
+      "loss": 0.8324,
+      "step": 7460
+    },
+    {
+      "epoch": 22.43,
+      "grad_norm": 20.358417510986328,
+      "learning_rate": 9.252252252252253e-06,
+      "loss": 0.8966,
+      "step": 7470
+    },
+    {
+      "epoch": 22.46,
+      "grad_norm": 15.353719711303711,
+      "learning_rate": 9.251251251251251e-06,
+      "loss": 0.8763,
+      "step": 7480
+    },
+    {
+      "epoch": 22.49,
+      "grad_norm": 16.366374969482422,
+      "learning_rate": 9.250250250250252e-06,
+      "loss": 0.8872,
+      "step": 7490
+    },
+    {
+      "epoch": 22.52,
+      "grad_norm": 20.56578254699707,
+      "learning_rate": 9.24924924924925e-06,
+      "loss": 0.8348,
+      "step": 7500
+    },
+    {
+      "epoch": 22.55,
+      "grad_norm": 14.54299259185791,
+      "learning_rate": 9.248248248248249e-06,
+      "loss": 0.8175,
+      "step": 7510
+    },
+    {
+      "epoch": 22.58,
+      "grad_norm": 25.86806869506836,
+      "learning_rate": 9.247247247247249e-06,
+      "loss": 0.8123,
+      "step": 7520
+    },
+    {
+      "epoch": 22.61,
+      "grad_norm": 14.958423614501953,
+      "learning_rate": 9.246246246246246e-06,
+      "loss": 0.8022,
+      "step": 7530
+    },
+    {
+      "epoch": 22.64,
+      "grad_norm": 15.241181373596191,
+      "learning_rate": 9.245245245245246e-06,
+      "loss": 0.8186,
+      "step": 7540
+    },
+    {
+      "epoch": 22.67,
+      "grad_norm": 13.112909317016602,
+      "learning_rate": 9.244244244244245e-06,
+      "loss": 0.8263,
+      "step": 7550
+    },
+    {
+      "epoch": 22.7,
+      "grad_norm": 17.80851173400879,
+      "learning_rate": 9.243243243243243e-06,
+      "loss": 0.875,
+      "step": 7560
+    },
+    {
+      "epoch": 22.73,
+      "grad_norm": 18.219003677368164,
+      "learning_rate": 9.242242242242244e-06,
+      "loss": 0.8504,
+      "step": 7570
+    },
+    {
+      "epoch": 22.76,
+      "grad_norm": 15.969429969787598,
+      "learning_rate": 9.241241241241242e-06,
+      "loss": 0.9018,
+      "step": 7580
+    },
+    {
+      "epoch": 22.79,
+      "grad_norm": 14.084677696228027,
+      "learning_rate": 9.240240240240241e-06,
+      "loss": 0.8464,
+      "step": 7590
+    },
+    {
+      "epoch": 22.82,
+      "grad_norm": 12.595839500427246,
+      "learning_rate": 9.239239239239241e-06,
+      "loss": 0.8395,
+      "step": 7600
+    },
+    {
+      "epoch": 22.85,
+      "grad_norm": 14.709906578063965,
+      "learning_rate": 9.238238238238238e-06,
+      "loss": 0.9069,
+      "step": 7610
+    },
+    {
+      "epoch": 22.88,
+      "grad_norm": 15.314925193786621,
+      "learning_rate": 9.237237237237238e-06,
+      "loss": 0.8889,
+      "step": 7620
+    },
+    {
+      "epoch": 22.91,
+      "grad_norm": 16.0838680267334,
+      "learning_rate": 9.236236236236237e-06,
+      "loss": 0.8378,
+      "step": 7630
+    },
+    {
+      "epoch": 22.94,
+      "grad_norm": 22.73562240600586,
+      "learning_rate": 9.235235235235236e-06,
+      "loss": 0.8536,
+      "step": 7640
+    },
+    {
+      "epoch": 22.97,
+      "grad_norm": 15.428640365600586,
+      "learning_rate": 9.234234234234236e-06,
+      "loss": 0.869,
+      "step": 7650
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.8528,
+      "eval_loss": 0.4703277349472046,
+      "eval_runtime": 12.7652,
+      "eval_samples_per_second": 783.377,
+      "eval_steps_per_second": 3.134,
+      "step": 7659
+    },
+    {
+      "epoch": 23.0,
+      "grad_norm": 15.16240119934082,
+      "learning_rate": 9.233233233233234e-06,
+      "loss": 0.7846,
+      "step": 7660
+    },
+    {
+      "epoch": 23.03,
+      "grad_norm": 19.369976043701172,
+      "learning_rate": 9.232232232232233e-06,
+      "loss": 0.8101,
+      "step": 7670
+    },
+    {
+      "epoch": 23.06,
+      "grad_norm": 17.12249183654785,
+      "learning_rate": 9.231231231231232e-06,
+      "loss": 0.8754,
+      "step": 7680
+    },
+    {
+      "epoch": 23.09,
+      "grad_norm": 15.322559356689453,
+      "learning_rate": 9.23023023023023e-06,
+      "loss": 0.8196,
+      "step": 7690
+    },
+    {
+      "epoch": 23.12,
+      "grad_norm": 17.422138214111328,
+      "learning_rate": 9.229229229229229e-06,
+      "loss": 0.7734,
+      "step": 7700
+    },
+    {
+      "epoch": 23.15,
+      "grad_norm": 14.549483299255371,
+      "learning_rate": 9.228228228228229e-06,
+      "loss": 0.8335,
+      "step": 7710
+    },
+    {
+      "epoch": 23.18,
+      "grad_norm": 14.285603523254395,
+      "learning_rate": 9.227227227227228e-06,
+      "loss": 0.9079,
+      "step": 7720
+    },
+    {
+      "epoch": 23.21,
+      "grad_norm": 15.609331130981445,
+      "learning_rate": 9.226226226226226e-06,
+      "loss": 0.7925,
+      "step": 7730
+    },
+    {
+      "epoch": 23.24,
+      "grad_norm": 15.03654670715332,
+      "learning_rate": 9.225225225225227e-06,
+      "loss": 0.8232,
+      "step": 7740
+    },
+    {
+      "epoch": 23.27,
+      "grad_norm": 15.817977905273438,
+      "learning_rate": 9.224224224224225e-06,
+      "loss": 0.8593,
+      "step": 7750
+    },
+    {
+      "epoch": 23.3,
+      "grad_norm": 17.551942825317383,
+      "learning_rate": 9.223223223223224e-06,
+      "loss": 0.8224,
+      "step": 7760
+    },
+    {
+      "epoch": 23.33,
+      "grad_norm": 15.3309907913208,
+      "learning_rate": 9.222222222222224e-06,
+      "loss": 0.8204,
+      "step": 7770
+    },
+    {
+      "epoch": 23.36,
+      "grad_norm": 18.83182144165039,
+      "learning_rate": 9.221221221221221e-06,
+      "loss": 0.8573,
+      "step": 7780
+    },
+    {
+      "epoch": 23.39,
+      "grad_norm": 18.249195098876953,
+      "learning_rate": 9.220220220220221e-06,
+      "loss": 0.7706,
+      "step": 7790
+    },
+    {
+      "epoch": 23.42,
+      "grad_norm": 17.53611946105957,
+      "learning_rate": 9.21921921921922e-06,
+      "loss": 0.8571,
+      "step": 7800
+    },
+    {
+      "epoch": 23.45,
+      "grad_norm": 18.392507553100586,
+      "learning_rate": 9.218218218218218e-06,
+      "loss": 0.7857,
+      "step": 7810
+    },
+    {
+      "epoch": 23.48,
+      "grad_norm": 16.387056350708008,
+      "learning_rate": 9.217217217217219e-06,
+      "loss": 0.7913,
+      "step": 7820
+    },
+    {
+      "epoch": 23.51,
+      "grad_norm": 14.469799995422363,
+      "learning_rate": 9.216216216216217e-06,
+      "loss": 0.8383,
+      "step": 7830
+    },
+    {
+      "epoch": 23.54,
+      "grad_norm": 15.943617820739746,
+      "learning_rate": 9.215215215215216e-06,
+      "loss": 0.782,
+      "step": 7840
+    },
+    {
+      "epoch": 23.57,
+      "grad_norm": 21.93720054626465,
+      "learning_rate": 9.214214214214216e-06,
+      "loss": 0.8644,
+      "step": 7850
+    },
+    {
+      "epoch": 23.6,
+      "grad_norm": 16.22055435180664,
+      "learning_rate": 9.213213213213213e-06,
+      "loss": 0.7544,
+      "step": 7860
+    },
+    {
+      "epoch": 23.63,
+      "grad_norm": 15.785012245178223,
+      "learning_rate": 9.212212212212213e-06,
+      "loss": 0.7947,
+      "step": 7870
+    },
+    {
+      "epoch": 23.66,
+      "grad_norm": 14.264601707458496,
+      "learning_rate": 9.211211211211212e-06,
+      "loss": 0.8322,
+      "step": 7880
+    },
+    {
+      "epoch": 23.69,
+      "grad_norm": 13.659516334533691,
+      "learning_rate": 9.21021021021021e-06,
+      "loss": 0.8537,
+      "step": 7890
+    },
+    {
+      "epoch": 23.72,
+      "grad_norm": 21.310932159423828,
+      "learning_rate": 9.20920920920921e-06,
+      "loss": 0.8742,
+      "step": 7900
+    },
+    {
+      "epoch": 23.75,
+      "grad_norm": 15.599264144897461,
+      "learning_rate": 9.20820820820821e-06,
+      "loss": 0.822,
+      "step": 7910
+    },
+    {
+      "epoch": 23.78,
+      "grad_norm": 24.196393966674805,
+      "learning_rate": 9.207207207207208e-06,
+      "loss": 0.8751,
+      "step": 7920
+    },
+    {
+      "epoch": 23.81,
+      "grad_norm": 15.778040885925293,
+      "learning_rate": 9.206206206206207e-06,
+      "loss": 0.8686,
+      "step": 7930
+    },
+    {
+      "epoch": 23.84,
+      "grad_norm": 18.238977432250977,
+      "learning_rate": 9.205205205205205e-06,
+      "loss": 0.8613,
+      "step": 7940
+    },
+    {
+      "epoch": 23.87,
+      "grad_norm": 14.943485260009766,
+      "learning_rate": 9.204204204204204e-06,
+      "loss": 0.8041,
+      "step": 7950
+    },
+    {
+      "epoch": 23.9,
+      "grad_norm": 12.482244491577148,
+      "learning_rate": 9.203203203203204e-06,
+      "loss": 0.776,
+      "step": 7960
+    },
+    {
+      "epoch": 23.93,
+      "grad_norm": 16.469688415527344,
+      "learning_rate": 9.202202202202203e-06,
+      "loss": 0.8753,
+      "step": 7970
+    },
+    {
+      "epoch": 23.96,
+      "grad_norm": 15.951470375061035,
+      "learning_rate": 9.201201201201201e-06,
+      "loss": 0.8528,
+      "step": 7980
+    },
+    {
+      "epoch": 23.99,
+      "grad_norm": 19.05841827392578,
+      "learning_rate": 9.200200200200202e-06,
+      "loss": 0.8681,
+      "step": 7990
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.8513,
+      "eval_loss": 0.47983139753341675,
+      "eval_runtime": 12.8651,
+      "eval_samples_per_second": 777.298,
+      "eval_steps_per_second": 3.109,
+      "step": 7992
+    },
+    {
+      "epoch": 24.02,
+      "grad_norm": 13.12854290008545,
+      "learning_rate": 9.1991991991992e-06,
+      "loss": 0.7567,
+      "step": 8000
+    },
+    {
+      "epoch": 24.05,
+      "grad_norm": 13.093955039978027,
+      "learning_rate": 9.198198198198199e-06,
+      "loss": 0.7827,
+      "step": 8010
+    },
+    {
+      "epoch": 24.08,
+      "grad_norm": 17.031505584716797,
+      "learning_rate": 9.197197197197199e-06,
+      "loss": 0.8131,
+      "step": 8020
+    },
+    {
+      "epoch": 24.11,
+      "grad_norm": 15.323765754699707,
+      "learning_rate": 9.196196196196196e-06,
+      "loss": 0.7802,
+      "step": 8030
+    },
+    {
+      "epoch": 24.14,
+      "grad_norm": 14.65307903289795,
+      "learning_rate": 9.195195195195196e-06,
+      "loss": 0.8292,
+      "step": 8040
+    },
+    {
+      "epoch": 24.17,
+      "grad_norm": 20.415098190307617,
+      "learning_rate": 9.194194194194195e-06,
+      "loss": 0.8012,
+      "step": 8050
+    },
+    {
+      "epoch": 24.2,
+      "grad_norm": 16.29452133178711,
+      "learning_rate": 9.193193193193194e-06,
+      "loss": 0.7967,
+      "step": 8060
+    },
+    {
+      "epoch": 24.23,
+      "grad_norm": 18.950679779052734,
+      "learning_rate": 9.192192192192194e-06,
+      "loss": 0.8356,
+      "step": 8070
+    },
+    {
+      "epoch": 24.26,
+      "grad_norm": 18.14234161376953,
+      "learning_rate": 9.191191191191192e-06,
+      "loss": 0.8139,
+      "step": 8080
+    },
+    {
+      "epoch": 24.29,
+      "grad_norm": 13.386857986450195,
+      "learning_rate": 9.190190190190191e-06,
+      "loss": 0.8111,
+      "step": 8090
+    },
+    {
+      "epoch": 24.32,
+      "grad_norm": 16.188678741455078,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.8458,
+      "step": 8100
+    },
+    {
+      "epoch": 24.35,
+      "grad_norm": 18.754507064819336,
+      "learning_rate": 9.188188188188188e-06,
+      "loss": 0.8107,
+      "step": 8110
+    },
+    {
+      "epoch": 24.38,
+      "grad_norm": 15.322908401489258,
+      "learning_rate": 9.187187187187187e-06,
+      "loss": 0.8701,
+      "step": 8120
+    },
+    {
+      "epoch": 24.41,
+      "grad_norm": 12.777388572692871,
+      "learning_rate": 9.186186186186187e-06,
+      "loss": 0.8625,
+      "step": 8130
+    },
+    {
+      "epoch": 24.44,
+      "grad_norm": 15.782909393310547,
+      "learning_rate": 9.185185185185186e-06,
+      "loss": 0.8565,
+      "step": 8140
+    },
+    {
+      "epoch": 24.47,
+      "grad_norm": 12.911823272705078,
+      "learning_rate": 9.184184184184184e-06,
+      "loss": 0.8595,
+      "step": 8150
+    },
+    {
+      "epoch": 24.5,
+      "grad_norm": 15.34666633605957,
+      "learning_rate": 9.183183183183185e-06,
+      "loss": 0.8232,
+      "step": 8160
+    },
+    {
+      "epoch": 24.53,
+      "grad_norm": 14.44251537322998,
+      "learning_rate": 9.182182182182183e-06,
+      "loss": 0.7846,
+      "step": 8170
+    },
+    {
+      "epoch": 24.56,
+      "grad_norm": 14.62682819366455,
+      "learning_rate": 9.181181181181182e-06,
+      "loss": 0.8122,
+      "step": 8180
+    },
+    {
+      "epoch": 24.59,
+      "grad_norm": 11.911892890930176,
+      "learning_rate": 9.18018018018018e-06,
+      "loss": 0.8487,
+      "step": 8190
+    },
+    {
+      "epoch": 24.62,
+      "grad_norm": 17.382272720336914,
+      "learning_rate": 9.179179179179179e-06,
+      "loss": 0.7474,
+      "step": 8200
+    },
+    {
+      "epoch": 24.65,
+      "grad_norm": 13.493112564086914,
+      "learning_rate": 9.17817817817818e-06,
+      "loss": 0.8379,
+      "step": 8210
+    },
+    {
+      "epoch": 24.68,
+      "grad_norm": 13.13019847869873,
+      "learning_rate": 9.177177177177178e-06,
+      "loss": 0.788,
+      "step": 8220
+    },
+    {
+      "epoch": 24.71,
+      "grad_norm": 17.745746612548828,
+      "learning_rate": 9.176176176176176e-06,
+      "loss": 0.8313,
+      "step": 8230
+    },
+    {
+      "epoch": 24.74,
+      "grad_norm": 16.05807113647461,
+      "learning_rate": 9.175175175175177e-06,
+      "loss": 0.8015,
+      "step": 8240
+    },
+    {
+      "epoch": 24.77,
+      "grad_norm": 17.024206161499023,
+      "learning_rate": 9.174174174174175e-06,
+      "loss": 0.8383,
+      "step": 8250
+    },
+    {
+      "epoch": 24.8,
+      "grad_norm": 16.1260986328125,
+      "learning_rate": 9.173173173173174e-06,
+      "loss": 0.8483,
+      "step": 8260
+    },
+    {
+      "epoch": 24.83,
+      "grad_norm": 16.126651763916016,
+      "learning_rate": 9.172172172172172e-06,
+      "loss": 0.8526,
+      "step": 8270
+    },
+    {
+      "epoch": 24.86,
+      "grad_norm": 12.465509414672852,
+      "learning_rate": 9.171171171171171e-06,
+      "loss": 0.7497,
+      "step": 8280
+    },
+    {
+      "epoch": 24.89,
+      "grad_norm": 14.012617111206055,
+      "learning_rate": 9.170170170170171e-06,
+      "loss": 0.8264,
+      "step": 8290
+    },
+    {
+      "epoch": 24.92,
+      "grad_norm": 15.739836692810059,
+      "learning_rate": 9.16916916916917e-06,
+      "loss": 0.7831,
+      "step": 8300
+    },
+    {
+      "epoch": 24.95,
+      "grad_norm": 16.17227554321289,
+      "learning_rate": 9.168168168168169e-06,
+      "loss": 0.8209,
+      "step": 8310
+    },
+    {
+      "epoch": 24.98,
+      "grad_norm": 14.855598449707031,
+      "learning_rate": 9.167167167167169e-06,
+      "loss": 0.7693,
+      "step": 8320
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.8523,
+      "eval_loss": 0.4679972529411316,
+      "eval_runtime": 12.7295,
+      "eval_samples_per_second": 785.579,
+      "eval_steps_per_second": 3.142,
+      "step": 8325
+    },
+    {
+      "epoch": 25.02,
+      "grad_norm": 19.4244384765625,
+      "learning_rate": 9.166166166166167e-06,
+      "loss": 0.8131,
+      "step": 8330
+    },
+    {
+      "epoch": 25.05,
+      "grad_norm": 17.67390251159668,
+      "learning_rate": 9.165165165165166e-06,
+      "loss": 0.7675,
+      "step": 8340
+    },
+    {
+      "epoch": 25.08,
+      "grad_norm": 16.680253982543945,
+      "learning_rate": 9.164164164164165e-06,
+      "loss": 0.8094,
+      "step": 8350
+    },
+    {
+      "epoch": 25.11,
+      "grad_norm": 19.503707885742188,
+      "learning_rate": 9.163163163163163e-06,
+      "loss": 0.8312,
+      "step": 8360
+    },
+    {
+      "epoch": 25.14,
+      "grad_norm": 19.490549087524414,
+      "learning_rate": 9.162162162162162e-06,
+      "loss": 0.8129,
+      "step": 8370
+    },
+    {
+      "epoch": 25.17,
+      "grad_norm": 15.471096992492676,
+      "learning_rate": 9.161161161161162e-06,
+      "loss": 0.8767,
+      "step": 8380
+    },
+    {
+      "epoch": 25.2,
+      "grad_norm": 16.655948638916016,
+      "learning_rate": 9.16016016016016e-06,
+      "loss": 0.8074,
+      "step": 8390
+    },
+    {
+      "epoch": 25.23,
+      "grad_norm": 24.472150802612305,
+      "learning_rate": 9.15915915915916e-06,
+      "loss": 0.6905,
+      "step": 8400
+    },
+    {
+      "epoch": 25.26,
+      "grad_norm": 12.948098182678223,
+      "learning_rate": 9.15815815815816e-06,
+      "loss": 0.7783,
+      "step": 8410
+    },
+    {
+      "epoch": 25.29,
+      "grad_norm": 20.246353149414062,
+      "learning_rate": 9.157157157157158e-06,
+      "loss": 0.7865,
+      "step": 8420
+    },
+    {
+      "epoch": 25.32,
+      "grad_norm": 16.572525024414062,
+      "learning_rate": 9.156156156156157e-06,
+      "loss": 0.8671,
+      "step": 8430
+    },
+    {
+      "epoch": 25.35,
+      "grad_norm": 13.740578651428223,
+      "learning_rate": 9.155155155155155e-06,
+      "loss": 0.7599,
+      "step": 8440
+    },
+    {
+      "epoch": 25.38,
+      "grad_norm": 19.841176986694336,
+      "learning_rate": 9.154154154154154e-06,
+      "loss": 0.8961,
+      "step": 8450
+    },
+    {
+      "epoch": 25.41,
+      "grad_norm": 14.382325172424316,
+      "learning_rate": 9.153153153153154e-06,
+      "loss": 0.8425,
+      "step": 8460
+    },
+    {
+      "epoch": 25.44,
+      "grad_norm": 16.1561336517334,
+      "learning_rate": 9.152152152152153e-06,
+      "loss": 0.8186,
+      "step": 8470
+    },
+    {
+      "epoch": 25.47,
+      "grad_norm": 17.0059871673584,
+      "learning_rate": 9.151151151151151e-06,
+      "loss": 0.746,
+      "step": 8480
+    },
+    {
+      "epoch": 25.5,
+      "grad_norm": 15.718467712402344,
+      "learning_rate": 9.150150150150152e-06,
+      "loss": 0.8708,
+      "step": 8490
+    },
+    {
+      "epoch": 25.53,
+      "grad_norm": 16.485553741455078,
+      "learning_rate": 9.14914914914915e-06,
+      "loss": 0.8399,
+      "step": 8500
+    },
+    {
+      "epoch": 25.56,
+      "grad_norm": 14.97546672821045,
+      "learning_rate": 9.148148148148149e-06,
+      "loss": 0.7853,
+      "step": 8510
+    },
+    {
+      "epoch": 25.59,
+      "grad_norm": 15.825803756713867,
+      "learning_rate": 9.147147147147147e-06,
+      "loss": 0.7837,
+      "step": 8520
+    },
+    {
+      "epoch": 25.62,
+      "grad_norm": 17.98794174194336,
+      "learning_rate": 9.146146146146146e-06,
+      "loss": 0.8777,
+      "step": 8530
+    },
+    {
+      "epoch": 25.65,
+      "grad_norm": 16.677520751953125,
+      "learning_rate": 9.145145145145146e-06,
+      "loss": 0.8902,
+      "step": 8540
+    },
+    {
+      "epoch": 25.68,
+      "grad_norm": 18.61178970336914,
+      "learning_rate": 9.144144144144145e-06,
+      "loss": 0.8479,
+      "step": 8550
+    },
+    {
+      "epoch": 25.71,
+      "grad_norm": 15.579206466674805,
+      "learning_rate": 9.143143143143144e-06,
+      "loss": 0.8276,
+      "step": 8560
+    },
+    {
+      "epoch": 25.74,
+      "grad_norm": 20.39088249206543,
+      "learning_rate": 9.142142142142144e-06,
+      "loss": 0.8162,
+      "step": 8570
+    },
+    {
+      "epoch": 25.77,
+      "grad_norm": 17.708942413330078,
+      "learning_rate": 9.141141141141142e-06,
+      "loss": 0.8538,
+      "step": 8580
+    },
+    {
+      "epoch": 25.8,
+      "grad_norm": 23.83163070678711,
+      "learning_rate": 9.140140140140141e-06,
+      "loss": 0.8433,
+      "step": 8590
+    },
+    {
+      "epoch": 25.83,
+      "grad_norm": 17.273645401000977,
+      "learning_rate": 9.13913913913914e-06,
+      "loss": 0.8071,
+      "step": 8600
+    },
+    {
+      "epoch": 25.86,
+      "grad_norm": 14.331090927124023,
+      "learning_rate": 9.138138138138138e-06,
+      "loss": 0.8062,
+      "step": 8610
+    },
+    {
+      "epoch": 25.89,
+      "grad_norm": 13.921991348266602,
+      "learning_rate": 9.137137137137137e-06,
+      "loss": 0.8676,
+      "step": 8620
+    },
+    {
+      "epoch": 25.92,
+      "grad_norm": 12.722929954528809,
+      "learning_rate": 9.136136136136137e-06,
+      "loss": 0.8033,
+      "step": 8630
+    },
+    {
+      "epoch": 25.95,
+      "grad_norm": 20.208118438720703,
+      "learning_rate": 9.135135135135136e-06,
+      "loss": 0.8174,
+      "step": 8640
+    },
+    {
+      "epoch": 25.98,
+      "grad_norm": 20.171621322631836,
+      "learning_rate": 9.134134134134134e-06,
+      "loss": 0.8693,
+      "step": 8650
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.8579,
+      "eval_loss": 0.46457383036613464,
+      "eval_runtime": 12.908,
+      "eval_samples_per_second": 774.716,
+      "eval_steps_per_second": 3.099,
+      "step": 8658
+    },
+    {
+      "epoch": 26.01,
+      "grad_norm": 13.121548652648926,
+      "learning_rate": 9.133133133133135e-06,
+      "loss": 0.7968,
+      "step": 8660
+    },
+    {
+      "epoch": 26.04,
+      "grad_norm": 14.341508865356445,
+      "learning_rate": 9.132132132132133e-06,
+      "loss": 0.797,
+      "step": 8670
+    },
+    {
+      "epoch": 26.07,
+      "grad_norm": 15.286103248596191,
+      "learning_rate": 9.131131131131132e-06,
+      "loss": 0.7673,
+      "step": 8680
+    },
+    {
+      "epoch": 26.1,
+      "grad_norm": 14.079110145568848,
+      "learning_rate": 9.13013013013013e-06,
+      "loss": 0.7776,
+      "step": 8690
+    },
+    {
+      "epoch": 26.13,
+      "grad_norm": 15.827554702758789,
+      "learning_rate": 9.129129129129129e-06,
+      "loss": 0.7723,
+      "step": 8700
+    },
+    {
+      "epoch": 26.16,
+      "grad_norm": 14.014419555664062,
+      "learning_rate": 9.12812812812813e-06,
+      "loss": 0.787,
+      "step": 8710
+    },
+    {
+      "epoch": 26.19,
+      "grad_norm": 18.37980842590332,
+      "learning_rate": 9.127127127127128e-06,
+      "loss": 0.8254,
+      "step": 8720
+    },
+    {
+      "epoch": 26.22,
+      "grad_norm": 15.558728218078613,
+      "learning_rate": 9.126126126126126e-06,
+      "loss": 0.7796,
+      "step": 8730
+    },
+    {
+      "epoch": 26.25,
+      "grad_norm": 20.04802894592285,
+      "learning_rate": 9.125125125125127e-06,
+      "loss": 0.7613,
+      "step": 8740
+    },
+    {
+      "epoch": 26.28,
+      "grad_norm": 19.22740936279297,
+      "learning_rate": 9.124124124124125e-06,
+      "loss": 0.7476,
+      "step": 8750
+    },
+    {
+      "epoch": 26.31,
+      "grad_norm": 15.683467864990234,
+      "learning_rate": 9.123123123123124e-06,
+      "loss": 0.8745,
+      "step": 8760
+    },
+    {
+      "epoch": 26.34,
+      "grad_norm": 15.375941276550293,
+      "learning_rate": 9.122122122122123e-06,
+      "loss": 0.7979,
+      "step": 8770
+    },
+    {
+      "epoch": 26.37,
+      "grad_norm": 23.70912742614746,
+      "learning_rate": 9.121121121121121e-06,
+      "loss": 0.8757,
+      "step": 8780
+    },
+    {
+      "epoch": 26.4,
+      "grad_norm": 16.594446182250977,
+      "learning_rate": 9.120120120120121e-06,
+      "loss": 0.8381,
+      "step": 8790
+    },
+    {
+      "epoch": 26.43,
+      "grad_norm": 20.395078659057617,
+      "learning_rate": 9.11911911911912e-06,
+      "loss": 0.8804,
+      "step": 8800
+    },
+    {
+      "epoch": 26.46,
+      "grad_norm": 15.868534088134766,
+      "learning_rate": 9.118118118118119e-06,
+      "loss": 0.8761,
+      "step": 8810
+    },
+    {
+      "epoch": 26.49,
+      "grad_norm": 18.077486038208008,
+      "learning_rate": 9.117117117117117e-06,
+      "loss": 0.8425,
+      "step": 8820
+    },
+    {
+      "epoch": 26.52,
+      "grad_norm": 15.2352294921875,
+      "learning_rate": 9.116116116116117e-06,
+      "loss": 0.7712,
+      "step": 8830
+    },
+    {
+      "epoch": 26.55,
+      "grad_norm": 14.5189208984375,
+      "learning_rate": 9.115115115115116e-06,
+      "loss": 0.7817,
+      "step": 8840
+    },
+    {
+      "epoch": 26.58,
+      "grad_norm": 15.7669038772583,
+      "learning_rate": 9.114114114114115e-06,
+      "loss": 0.8065,
+      "step": 8850
+    },
+    {
+      "epoch": 26.61,
+      "grad_norm": 16.65861701965332,
+      "learning_rate": 9.113113113113113e-06,
+      "loss": 0.8495,
+      "step": 8860
+    },
+    {
+      "epoch": 26.64,
+      "grad_norm": 14.653789520263672,
+      "learning_rate": 9.112112112112112e-06,
+      "loss": 0.8059,
+      "step": 8870
+    },
+    {
+      "epoch": 26.67,
+      "grad_norm": 25.060998916625977,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 0.8203,
+      "step": 8880
+    },
+    {
+      "epoch": 26.7,
+      "grad_norm": 10.723742485046387,
+      "learning_rate": 9.11011011011011e-06,
+      "loss": 0.7768,
+      "step": 8890
+    },
+    {
+      "epoch": 26.73,
+      "grad_norm": 16.3524227142334,
+      "learning_rate": 9.10910910910911e-06,
+      "loss": 0.7948,
+      "step": 8900
+    },
+    {
+      "epoch": 26.76,
+      "grad_norm": 13.840607643127441,
+      "learning_rate": 9.10810810810811e-06,
+      "loss": 0.7554,
+      "step": 8910
+    },
+    {
+      "epoch": 26.79,
+      "grad_norm": 15.388121604919434,
+      "learning_rate": 9.107107107107108e-06,
+      "loss": 0.8108,
+      "step": 8920
+    },
+    {
+      "epoch": 26.82,
+      "grad_norm": 15.419222831726074,
+      "learning_rate": 9.106106106106107e-06,
+      "loss": 0.8551,
+      "step": 8930
+    },
+    {
+      "epoch": 26.85,
+      "grad_norm": 13.759719848632812,
+      "learning_rate": 9.105105105105105e-06,
+      "loss": 0.7758,
+      "step": 8940
+    },
+    {
+      "epoch": 26.88,
+      "grad_norm": 20.2331600189209,
+      "learning_rate": 9.104104104104104e-06,
+      "loss": 0.7934,
+      "step": 8950
+    },
+    {
+      "epoch": 26.91,
+      "grad_norm": 13.752047538757324,
+      "learning_rate": 9.103103103103104e-06,
+      "loss": 0.7795,
+      "step": 8960
+    },
+    {
+      "epoch": 26.94,
+      "grad_norm": 20.42212677001953,
+      "learning_rate": 9.102102102102103e-06,
+      "loss": 0.842,
+      "step": 8970
+    },
+    {
+      "epoch": 26.97,
+      "grad_norm": 15.064266204833984,
+      "learning_rate": 9.101101101101101e-06,
+      "loss": 0.8582,
+      "step": 8980
+    },
+    {
+      "epoch": 27.0,
+      "grad_norm": 13.374460220336914,
+      "learning_rate": 9.100100100100102e-06,
+      "loss": 0.8041,
+      "step": 8990
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.8555,
+      "eval_loss": 0.46856260299682617,
+      "eval_runtime": 12.6948,
+      "eval_samples_per_second": 787.727,
+      "eval_steps_per_second": 3.151,
+      "step": 8991
+    },
+    {
+      "epoch": 27.03,
+      "grad_norm": 15.646354675292969,
+      "learning_rate": 9.0990990990991e-06,
+      "loss": 0.7201,
+      "step": 9000
+    },
+    {
+      "epoch": 27.06,
+      "grad_norm": 14.156930923461914,
+      "learning_rate": 9.098098098098099e-06,
+      "loss": 0.7586,
+      "step": 9010
+    },
+    {
+      "epoch": 27.09,
+      "grad_norm": 18.541675567626953,
+      "learning_rate": 9.097097097097098e-06,
+      "loss": 0.8411,
+      "step": 9020
+    },
+    {
+      "epoch": 27.12,
+      "grad_norm": 15.928442001342773,
+      "learning_rate": 9.096096096096096e-06,
+      "loss": 0.8087,
+      "step": 9030
+    },
+    {
+      "epoch": 27.15,
+      "grad_norm": 23.12171745300293,
+      "learning_rate": 9.095095095095095e-06,
+      "loss": 0.8369,
+      "step": 9040
+    },
+    {
+      "epoch": 27.18,
+      "grad_norm": 13.158681869506836,
+      "learning_rate": 9.094094094094095e-06,
+      "loss": 0.8279,
+      "step": 9050
+    },
+    {
+      "epoch": 27.21,
+      "grad_norm": 12.238262176513672,
+      "learning_rate": 9.093093093093094e-06,
+      "loss": 0.7691,
+      "step": 9060
+    },
+    {
+      "epoch": 27.24,
+      "grad_norm": 17.15790367126465,
+      "learning_rate": 9.092092092092092e-06,
+      "loss": 0.7774,
+      "step": 9070
+    },
+    {
+      "epoch": 27.27,
+      "grad_norm": 12.419235229492188,
+      "learning_rate": 9.091091091091093e-06,
+      "loss": 0.7231,
+      "step": 9080
+    },
+    {
+      "epoch": 27.3,
+      "grad_norm": 12.092683792114258,
+      "learning_rate": 9.090090090090091e-06,
+      "loss": 0.833,
+      "step": 9090
+    },
+    {
+      "epoch": 27.33,
+      "grad_norm": 15.581094741821289,
+      "learning_rate": 9.08908908908909e-06,
+      "loss": 0.7504,
+      "step": 9100
+    },
+    {
+      "epoch": 27.36,
+      "grad_norm": 18.712749481201172,
+      "learning_rate": 9.088088088088088e-06,
+      "loss": 0.8297,
+      "step": 9110
+    },
+    {
+      "epoch": 27.39,
+      "grad_norm": 16.576900482177734,
+      "learning_rate": 9.087087087087087e-06,
+      "loss": 0.8084,
+      "step": 9120
+    },
+    {
+      "epoch": 27.42,
+      "grad_norm": 28.18408966064453,
+      "learning_rate": 9.086086086086087e-06,
+      "loss": 0.8344,
+      "step": 9130
+    },
+    {
+      "epoch": 27.45,
+      "grad_norm": 18.83150291442871,
+      "learning_rate": 9.085085085085086e-06,
+      "loss": 0.792,
+      "step": 9140
+    },
+    {
+      "epoch": 27.48,
+      "grad_norm": 19.39872932434082,
+      "learning_rate": 9.084084084084084e-06,
+      "loss": 0.8539,
+      "step": 9150
+    },
+    {
+      "epoch": 27.51,
+      "grad_norm": 15.377179145812988,
+      "learning_rate": 9.083083083083085e-06,
+      "loss": 0.8331,
+      "step": 9160
+    },
+    {
+      "epoch": 27.54,
+      "grad_norm": 15.049494743347168,
+      "learning_rate": 9.082082082082083e-06,
+      "loss": 0.7915,
+      "step": 9170
+    },
+    {
+      "epoch": 27.57,
+      "grad_norm": 14.656966209411621,
+      "learning_rate": 9.081081081081082e-06,
+      "loss": 0.7344,
+      "step": 9180
+    },
+    {
+      "epoch": 27.6,
+      "grad_norm": 18.953798294067383,
+      "learning_rate": 9.08008008008008e-06,
+      "loss": 0.7675,
+      "step": 9190
+    },
+    {
+      "epoch": 27.63,
+      "grad_norm": 15.632416725158691,
+      "learning_rate": 9.079079079079079e-06,
+      "loss": 0.8395,
+      "step": 9200
+    },
+    {
+      "epoch": 27.66,
+      "grad_norm": 19.34852409362793,
+      "learning_rate": 9.07807807807808e-06,
+      "loss": 0.7695,
+      "step": 9210
+    },
+    {
+      "epoch": 27.69,
+      "grad_norm": 16.209014892578125,
+      "learning_rate": 9.077077077077078e-06,
+      "loss": 0.7672,
+      "step": 9220
+    },
+    {
+      "epoch": 27.72,
+      "grad_norm": 19.034046173095703,
+      "learning_rate": 9.076076076076077e-06,
+      "loss": 0.9088,
+      "step": 9230
+    },
+    {
+      "epoch": 27.75,
+      "grad_norm": 17.229251861572266,
+      "learning_rate": 9.075075075075077e-06,
+      "loss": 0.8759,
+      "step": 9240
+    },
+    {
+      "epoch": 27.78,
+      "grad_norm": 14.129185676574707,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 0.8082,
+      "step": 9250
+    },
+    {
+      "epoch": 27.81,
+      "grad_norm": 13.324420928955078,
+      "learning_rate": 9.073073073073074e-06,
+      "loss": 0.777,
+      "step": 9260
+    },
+    {
+      "epoch": 27.84,
+      "grad_norm": 16.472885131835938,
+      "learning_rate": 9.072072072072073e-06,
+      "loss": 0.809,
+      "step": 9270
+    },
+    {
+      "epoch": 27.87,
+      "grad_norm": 14.68637466430664,
+      "learning_rate": 9.071071071071071e-06,
+      "loss": 0.7496,
+      "step": 9280
+    },
+    {
+      "epoch": 27.9,
+      "grad_norm": 13.848377227783203,
+      "learning_rate": 9.07007007007007e-06,
+      "loss": 0.8411,
+      "step": 9290
+    },
+    {
+      "epoch": 27.93,
+      "grad_norm": 17.72799301147461,
+      "learning_rate": 9.06906906906907e-06,
+      "loss": 0.7486,
+      "step": 9300
+    },
+    {
+      "epoch": 27.96,
+      "grad_norm": 16.831594467163086,
+      "learning_rate": 9.068068068068069e-06,
+      "loss": 0.8027,
+      "step": 9310
+    },
+    {
+      "epoch": 27.99,
+      "grad_norm": 30.738222122192383,
+      "learning_rate": 9.067067067067067e-06,
+      "loss": 0.8036,
+      "step": 9320
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.8578,
+      "eval_loss": 0.46087294816970825,
+      "eval_runtime": 12.7214,
+      "eval_samples_per_second": 786.076,
+      "eval_steps_per_second": 3.144,
+      "step": 9324
+    },
+    {
+      "epoch": 28.02,
+      "grad_norm": 11.911028861999512,
+      "learning_rate": 9.066066066066068e-06,
+      "loss": 0.7672,
+      "step": 9330
+    },
+    {
+      "epoch": 28.05,
+      "grad_norm": 14.490723609924316,
+      "learning_rate": 9.065065065065066e-06,
+      "loss": 0.7613,
+      "step": 9340
+    },
+    {
+      "epoch": 28.08,
+      "grad_norm": 19.9595947265625,
+      "learning_rate": 9.064064064064065e-06,
+      "loss": 0.7991,
+      "step": 9350
+    },
+    {
+      "epoch": 28.11,
+      "grad_norm": 17.42326545715332,
+      "learning_rate": 9.063063063063063e-06,
+      "loss": 0.8108,
+      "step": 9360
+    },
+    {
+      "epoch": 28.14,
+      "grad_norm": 15.845438957214355,
+      "learning_rate": 9.062062062062062e-06,
+      "loss": 0.7008,
+      "step": 9370
+    },
+    {
+      "epoch": 28.17,
+      "grad_norm": 20.148286819458008,
+      "learning_rate": 9.061061061061062e-06,
+      "loss": 0.7418,
+      "step": 9380
+    },
+    {
+      "epoch": 28.2,
+      "grad_norm": 13.345690727233887,
+      "learning_rate": 9.06006006006006e-06,
+      "loss": 0.7152,
+      "step": 9390
+    },
+    {
+      "epoch": 28.23,
+      "grad_norm": 12.969523429870605,
+      "learning_rate": 9.05905905905906e-06,
+      "loss": 0.8165,
+      "step": 9400
+    },
+    {
+      "epoch": 28.26,
+      "grad_norm": 15.857499122619629,
+      "learning_rate": 9.05805805805806e-06,
+      "loss": 0.73,
+      "step": 9410
+    },
+    {
+      "epoch": 28.29,
+      "grad_norm": 14.983546257019043,
+      "learning_rate": 9.057057057057058e-06,
+      "loss": 0.8108,
+      "step": 9420
+    },
+    {
+      "epoch": 28.32,
+      "grad_norm": 16.10476303100586,
+      "learning_rate": 9.056056056056057e-06,
+      "loss": 0.7254,
+      "step": 9430
+    },
+    {
+      "epoch": 28.35,
+      "grad_norm": 12.998758316040039,
+      "learning_rate": 9.055055055055055e-06,
+      "loss": 0.7881,
+      "step": 9440
+    },
+    {
+      "epoch": 28.38,
+      "grad_norm": 17.046140670776367,
+      "learning_rate": 9.054054054054054e-06,
+      "loss": 0.8049,
+      "step": 9450
+    },
+    {
+      "epoch": 28.41,
+      "grad_norm": 15.55833625793457,
+      "learning_rate": 9.053053053053054e-06,
+      "loss": 0.7728,
+      "step": 9460
+    },
+    {
+      "epoch": 28.44,
+      "grad_norm": 16.298221588134766,
+      "learning_rate": 9.052052052052053e-06,
+      "loss": 0.794,
+      "step": 9470
+    },
+    {
+      "epoch": 28.47,
+      "grad_norm": 16.310712814331055,
+      "learning_rate": 9.051051051051052e-06,
+      "loss": 0.835,
+      "step": 9480
+    },
+    {
+      "epoch": 28.5,
+      "grad_norm": 15.452524185180664,
+      "learning_rate": 9.05005005005005e-06,
+      "loss": 0.8311,
+      "step": 9490
+    },
+    {
+      "epoch": 28.53,
+      "grad_norm": 13.287662506103516,
+      "learning_rate": 9.04904904904905e-06,
+      "loss": 0.791,
+      "step": 9500
+    },
+    {
+      "epoch": 28.56,
+      "grad_norm": 11.791207313537598,
+      "learning_rate": 9.048048048048049e-06,
+      "loss": 0.8211,
+      "step": 9510
+    },
+    {
+      "epoch": 28.59,
+      "grad_norm": 14.900876998901367,
+      "learning_rate": 9.047047047047048e-06,
+      "loss": 0.7735,
+      "step": 9520
+    },
+    {
+      "epoch": 28.62,
+      "grad_norm": 18.341766357421875,
+      "learning_rate": 9.046046046046046e-06,
+      "loss": 0.8503,
+      "step": 9530
+    },
+    {
+      "epoch": 28.65,
+      "grad_norm": 18.77469253540039,
+      "learning_rate": 9.045045045045045e-06,
+      "loss": 0.7304,
+      "step": 9540
+    },
+    {
+      "epoch": 28.68,
+      "grad_norm": 18.09326171875,
+      "learning_rate": 9.044044044044045e-06,
+      "loss": 0.7796,
+      "step": 9550
+    },
+    {
+      "epoch": 28.71,
+      "grad_norm": 16.77538299560547,
+      "learning_rate": 9.043043043043044e-06,
+      "loss": 0.8585,
+      "step": 9560
+    },
+    {
+      "epoch": 28.74,
+      "grad_norm": 14.218920707702637,
+      "learning_rate": 9.042042042042042e-06,
+      "loss": 0.8425,
+      "step": 9570
+    },
+    {
+      "epoch": 28.77,
+      "grad_norm": 15.359567642211914,
+      "learning_rate": 9.041041041041043e-06,
+      "loss": 0.7995,
+      "step": 9580
+    },
+    {
+      "epoch": 28.8,
+      "grad_norm": 17.3696231842041,
+      "learning_rate": 9.040040040040041e-06,
+      "loss": 0.8462,
+      "step": 9590
+    },
+    {
+      "epoch": 28.83,
+      "grad_norm": 17.48063850402832,
+      "learning_rate": 9.03903903903904e-06,
+      "loss": 0.794,
+      "step": 9600
+    },
+    {
+      "epoch": 28.86,
+      "grad_norm": 14.394133567810059,
+      "learning_rate": 9.038038038038038e-06,
+      "loss": 0.8378,
+      "step": 9610
+    },
+    {
+      "epoch": 28.89,
+      "grad_norm": 22.86805534362793,
+      "learning_rate": 9.037037037037037e-06,
+      "loss": 0.7753,
+      "step": 9620
+    },
+    {
+      "epoch": 28.92,
+      "grad_norm": 11.59033203125,
+      "learning_rate": 9.036036036036037e-06,
+      "loss": 0.7755,
+      "step": 9630
+    },
+    {
+      "epoch": 28.95,
+      "grad_norm": 21.878067016601562,
+      "learning_rate": 9.035035035035036e-06,
+      "loss": 0.843,
+      "step": 9640
+    },
+    {
+      "epoch": 28.98,
+      "grad_norm": 12.910211563110352,
+      "learning_rate": 9.034034034034034e-06,
+      "loss": 0.7571,
+      "step": 9650
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.8616,
+      "eval_loss": 0.4597092270851135,
+      "eval_runtime": 12.7071,
+      "eval_samples_per_second": 786.965,
+      "eval_steps_per_second": 3.148,
+      "step": 9657
+    },
+    {
+      "epoch": 29.01,
+      "grad_norm": 22.672311782836914,
+      "learning_rate": 9.033033033033035e-06,
+      "loss": 0.7749,
+      "step": 9660
+    },
+    {
+      "epoch": 29.04,
+      "grad_norm": 20.296001434326172,
+      "learning_rate": 9.032032032032033e-06,
+      "loss": 0.7856,
+      "step": 9670
+    },
+    {
+      "epoch": 29.07,
+      "grad_norm": 15.42845344543457,
+      "learning_rate": 9.031031031031032e-06,
+      "loss": 0.7632,
+      "step": 9680
+    },
+    {
+      "epoch": 29.1,
+      "grad_norm": 18.354835510253906,
+      "learning_rate": 9.03003003003003e-06,
+      "loss": 0.7764,
+      "step": 9690
+    },
+    {
+      "epoch": 29.13,
+      "grad_norm": 20.938955307006836,
+      "learning_rate": 9.029029029029029e-06,
+      "loss": 0.7339,
+      "step": 9700
+    },
+    {
+      "epoch": 29.16,
+      "grad_norm": 14.629878997802734,
+      "learning_rate": 9.02802802802803e-06,
+      "loss": 0.7615,
+      "step": 9710
+    },
+    {
+      "epoch": 29.19,
+      "grad_norm": 13.274352073669434,
+      "learning_rate": 9.027027027027028e-06,
+      "loss": 0.7745,
+      "step": 9720
+    },
+    {
+      "epoch": 29.22,
+      "grad_norm": 15.748404502868652,
+      "learning_rate": 9.026026026026027e-06,
+      "loss": 0.7845,
+      "step": 9730
+    },
+    {
+      "epoch": 29.25,
+      "grad_norm": 15.212640762329102,
+      "learning_rate": 9.025025025025025e-06,
+      "loss": 0.7967,
+      "step": 9740
+    },
+    {
+      "epoch": 29.28,
+      "grad_norm": 17.692638397216797,
+      "learning_rate": 9.024024024024025e-06,
+      "loss": 0.7447,
+      "step": 9750
+    },
+    {
+      "epoch": 29.31,
+      "grad_norm": 15.170554161071777,
+      "learning_rate": 9.023023023023024e-06,
+      "loss": 0.7363,
+      "step": 9760
+    },
+    {
+      "epoch": 29.34,
+      "grad_norm": 14.743155479431152,
+      "learning_rate": 9.022022022022023e-06,
+      "loss": 0.7615,
+      "step": 9770
+    },
+    {
+      "epoch": 29.37,
+      "grad_norm": 18.1495361328125,
+      "learning_rate": 9.021021021021021e-06,
+      "loss": 0.7413,
+      "step": 9780
+    },
+    {
+      "epoch": 29.4,
+      "grad_norm": 19.55647087097168,
+      "learning_rate": 9.02002002002002e-06,
+      "loss": 0.7874,
+      "step": 9790
+    },
+    {
+      "epoch": 29.43,
+      "grad_norm": 13.923855781555176,
+      "learning_rate": 9.01901901901902e-06,
+      "loss": 0.7791,
+      "step": 9800
+    },
+    {
+      "epoch": 29.46,
+      "grad_norm": 14.469110488891602,
+      "learning_rate": 9.018018018018019e-06,
+      "loss": 0.7539,
+      "step": 9810
+    },
+    {
+      "epoch": 29.49,
+      "grad_norm": 18.734216690063477,
+      "learning_rate": 9.017017017017017e-06,
+      "loss": 0.7518,
+      "step": 9820
+    },
+    {
+      "epoch": 29.52,
+      "grad_norm": 29.21765899658203,
+      "learning_rate": 9.016016016016018e-06,
+      "loss": 0.788,
+      "step": 9830
+    },
+    {
+      "epoch": 29.55,
+      "grad_norm": 18.254535675048828,
+      "learning_rate": 9.015015015015016e-06,
+      "loss": 0.8137,
+      "step": 9840
+    },
+    {
+      "epoch": 29.58,
+      "grad_norm": 24.3439998626709,
+      "learning_rate": 9.014014014014015e-06,
+      "loss": 0.7546,
+      "step": 9850
+    },
+    {
+      "epoch": 29.61,
+      "grad_norm": 21.63543701171875,
+      "learning_rate": 9.013013013013013e-06,
+      "loss": 0.8033,
+      "step": 9860
+    },
+    {
+      "epoch": 29.64,
+      "grad_norm": 20.718181610107422,
+      "learning_rate": 9.012012012012012e-06,
+      "loss": 0.7559,
+      "step": 9870
+    },
+    {
+      "epoch": 29.67,
+      "grad_norm": 18.576004028320312,
+      "learning_rate": 9.011011011011012e-06,
+      "loss": 0.6953,
+      "step": 9880
+    },
+    {
+      "epoch": 29.7,
+      "grad_norm": 13.842483520507812,
+      "learning_rate": 9.010010010010011e-06,
+      "loss": 0.8485,
+      "step": 9890
+    },
+    {
+      "epoch": 29.73,
+      "grad_norm": 12.415961265563965,
+      "learning_rate": 9.00900900900901e-06,
+      "loss": 0.7197,
+      "step": 9900
+    },
+    {
+      "epoch": 29.76,
+      "grad_norm": 21.572906494140625,
+      "learning_rate": 9.00800800800801e-06,
+      "loss": 0.7952,
+      "step": 9910
+    },
+    {
+      "epoch": 29.79,
+      "grad_norm": 16.539274215698242,
+      "learning_rate": 9.007007007007008e-06,
+      "loss": 0.8305,
+      "step": 9920
+    },
+    {
+      "epoch": 29.82,
+      "grad_norm": 24.245912551879883,
+      "learning_rate": 9.006006006006007e-06,
+      "loss": 0.7945,
+      "step": 9930
+    },
+    {
+      "epoch": 29.85,
+      "grad_norm": 16.62843132019043,
+      "learning_rate": 9.005005005005006e-06,
+      "loss": 0.8332,
+      "step": 9940
+    },
+    {
+      "epoch": 29.88,
+      "grad_norm": 17.65282440185547,
+      "learning_rate": 9.004004004004004e-06,
+      "loss": 0.7788,
+      "step": 9950
+    },
+    {
+      "epoch": 29.91,
+      "grad_norm": 14.350862503051758,
+      "learning_rate": 9.003003003003003e-06,
+      "loss": 0.8164,
+      "step": 9960
+    },
+    {
+      "epoch": 29.94,
+      "grad_norm": 14.50328540802002,
+      "learning_rate": 9.002002002002003e-06,
+      "loss": 0.7701,
+      "step": 9970
+    },
+    {
+      "epoch": 29.97,
+      "grad_norm": 15.73355484008789,
+      "learning_rate": 9.001001001001002e-06,
+      "loss": 0.7745,
+      "step": 9980
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 62.948577880859375,
+      "learning_rate": 9e-06,
+      "loss": 0.7666,
+      "step": 9990
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.8606,
+      "eval_loss": 0.45814234018325806,
+      "eval_runtime": 12.6385,
+      "eval_samples_per_second": 791.23,
+      "eval_steps_per_second": 3.165,
+      "step": 9990
+    },
+    {
+      "epoch": 30.03,
+      "grad_norm": 16.523704528808594,
+      "learning_rate": 8.998998998999e-06,
+      "loss": 0.8094,
+      "step": 10000
+    },
+    {
+      "epoch": 30.06,
+      "grad_norm": 12.528546333312988,
+      "learning_rate": 8.997997997997999e-06,
+      "loss": 0.8251,
+      "step": 10010
+    },
+    {
+      "epoch": 30.09,
+      "grad_norm": 16.390399932861328,
+      "learning_rate": 8.996996996996998e-06,
+      "loss": 0.7715,
+      "step": 10020
+    },
+    {
+      "epoch": 30.12,
+      "grad_norm": 17.987390518188477,
+      "learning_rate": 8.995995995995996e-06,
+      "loss": 0.7799,
+      "step": 10030
+    },
+    {
+      "epoch": 30.15,
+      "grad_norm": 17.51533317565918,
+      "learning_rate": 8.994994994994995e-06,
+      "loss": 0.7342,
+      "step": 10040
+    },
+    {
+      "epoch": 30.18,
+      "grad_norm": 13.23087215423584,
+      "learning_rate": 8.993993993993995e-06,
+      "loss": 0.691,
+      "step": 10050
+    },
+    {
+      "epoch": 30.21,
+      "grad_norm": 16.02936553955078,
+      "learning_rate": 8.992992992992994e-06,
+      "loss": 0.7646,
+      "step": 10060
+    },
+    {
+      "epoch": 30.24,
+      "grad_norm": 13.150365829467773,
+      "learning_rate": 8.991991991991992e-06,
+      "loss": 0.6925,
+      "step": 10070
+    },
+    {
+      "epoch": 30.27,
+      "grad_norm": 20.552528381347656,
+      "learning_rate": 8.990990990990993e-06,
+      "loss": 0.7867,
+      "step": 10080
+    },
+    {
+      "epoch": 30.3,
+      "grad_norm": 14.539362907409668,
+      "learning_rate": 8.989989989989991e-06,
+      "loss": 0.691,
+      "step": 10090
+    },
+    {
+      "epoch": 30.33,
+      "grad_norm": 18.682815551757812,
+      "learning_rate": 8.98898898898899e-06,
+      "loss": 0.7358,
+      "step": 10100
+    },
+    {
+      "epoch": 30.36,
+      "grad_norm": 25.57851219177246,
+      "learning_rate": 8.987987987987988e-06,
+      "loss": 0.6709,
+      "step": 10110
+    },
+    {
+      "epoch": 30.39,
+      "grad_norm": 23.194116592407227,
+      "learning_rate": 8.986986986986987e-06,
+      "loss": 0.7437,
+      "step": 10120
+    },
+    {
+      "epoch": 30.42,
+      "grad_norm": 23.166088104248047,
+      "learning_rate": 8.985985985985987e-06,
+      "loss": 0.7844,
+      "step": 10130
+    },
+    {
+      "epoch": 30.45,
+      "grad_norm": 15.125076293945312,
+      "learning_rate": 8.984984984984986e-06,
+      "loss": 0.7677,
+      "step": 10140
+    },
+    {
+      "epoch": 30.48,
+      "grad_norm": 16.771602630615234,
+      "learning_rate": 8.983983983983985e-06,
+      "loss": 0.7418,
+      "step": 10150
+    },
+    {
+      "epoch": 30.51,
+      "grad_norm": 15.16159725189209,
+      "learning_rate": 8.982982982982985e-06,
+      "loss": 0.7895,
+      "step": 10160
+    },
+    {
+      "epoch": 30.54,
+      "grad_norm": 14.736344337463379,
+      "learning_rate": 8.981981981981983e-06,
+      "loss": 0.7441,
+      "step": 10170
+    },
+    {
+      "epoch": 30.57,
+      "grad_norm": 17.855131149291992,
+      "learning_rate": 8.980980980980982e-06,
+      "loss": 0.6948,
+      "step": 10180
+    },
+    {
+      "epoch": 30.6,
+      "grad_norm": 16.532987594604492,
+      "learning_rate": 8.97997997997998e-06,
+      "loss": 0.743,
+      "step": 10190
+    },
+    {
+      "epoch": 30.63,
+      "grad_norm": 16.70667266845703,
+      "learning_rate": 8.97897897897898e-06,
+      "loss": 0.7552,
+      "step": 10200
+    },
+    {
+      "epoch": 30.66,
+      "grad_norm": 15.976407051086426,
+      "learning_rate": 8.977977977977978e-06,
+      "loss": 0.709,
+      "step": 10210
+    },
+    {
+      "epoch": 30.69,
+      "grad_norm": 24.74825096130371,
+      "learning_rate": 8.976976976976978e-06,
+      "loss": 0.7439,
+      "step": 10220
+    },
+    {
+      "epoch": 30.72,
+      "grad_norm": 26.14236068725586,
+      "learning_rate": 8.975975975975977e-06,
+      "loss": 0.7512,
+      "step": 10230
+    },
+    {
+      "epoch": 30.75,
+      "grad_norm": 15.58338737487793,
+      "learning_rate": 8.974974974974975e-06,
+      "loss": 0.7207,
+      "step": 10240
+    },
+    {
+      "epoch": 30.78,
+      "grad_norm": 13.44525146484375,
+      "learning_rate": 8.973973973973976e-06,
+      "loss": 0.775,
+      "step": 10250
+    },
+    {
+      "epoch": 30.81,
+      "grad_norm": 12.716086387634277,
+      "learning_rate": 8.972972972972974e-06,
+      "loss": 0.7855,
+      "step": 10260
+    },
+    {
+      "epoch": 30.84,
+      "grad_norm": 17.73958396911621,
+      "learning_rate": 8.971971971971973e-06,
+      "loss": 0.7441,
+      "step": 10270
+    },
+    {
+      "epoch": 30.87,
+      "grad_norm": 13.41761589050293,
+      "learning_rate": 8.970970970970971e-06,
+      "loss": 0.7244,
+      "step": 10280
+    },
+    {
+      "epoch": 30.9,
+      "grad_norm": 15.482220649719238,
+      "learning_rate": 8.96996996996997e-06,
+      "loss": 0.7412,
+      "step": 10290
+    },
+    {
+      "epoch": 30.93,
+      "grad_norm": 21.90937614440918,
+      "learning_rate": 8.96896896896897e-06,
+      "loss": 0.784,
+      "step": 10300
+    },
+    {
+      "epoch": 30.96,
+      "grad_norm": 18.506118774414062,
+      "learning_rate": 8.967967967967969e-06,
+      "loss": 0.7994,
+      "step": 10310
+    },
+    {
+      "epoch": 30.99,
+      "grad_norm": 16.833866119384766,
+      "learning_rate": 8.966966966966967e-06,
+      "loss": 0.7226,
+      "step": 10320
+    },
+    {
+      "epoch": 31.0,
+      "eval_accuracy": 0.8601,
+      "eval_loss": 0.45692509412765503,
+      "eval_runtime": 12.6373,
+      "eval_samples_per_second": 791.306,
+      "eval_steps_per_second": 3.165,
+      "step": 10323
+    },
+    {
+      "epoch": 31.02,
+      "grad_norm": 19.576566696166992,
+      "learning_rate": 8.965965965965968e-06,
+      "loss": 0.7177,
+      "step": 10330
+    },
+    {
+      "epoch": 31.05,
+      "grad_norm": 15.05361557006836,
+      "learning_rate": 8.964964964964966e-06,
+      "loss": 0.8626,
+      "step": 10340
+    },
+    {
+      "epoch": 31.08,
+      "grad_norm": 15.818574905395508,
+      "learning_rate": 8.963963963963965e-06,
+      "loss": 0.817,
+      "step": 10350
+    },
+    {
+      "epoch": 31.11,
+      "grad_norm": 15.349098205566406,
+      "learning_rate": 8.962962962962963e-06,
+      "loss": 0.7074,
+      "step": 10360
+    },
+    {
+      "epoch": 31.14,
+      "grad_norm": 17.57956314086914,
+      "learning_rate": 8.961961961961962e-06,
+      "loss": 0.6718,
+      "step": 10370
+    },
+    {
+      "epoch": 31.17,
+      "grad_norm": 17.0306453704834,
+      "learning_rate": 8.960960960960962e-06,
+      "loss": 0.7001,
+      "step": 10380
+    },
+    {
+      "epoch": 31.2,
+      "grad_norm": 19.56716537475586,
+      "learning_rate": 8.959959959959961e-06,
+      "loss": 0.7383,
+      "step": 10390
+    },
+    {
+      "epoch": 31.23,
+      "grad_norm": 14.34196949005127,
+      "learning_rate": 8.95895895895896e-06,
+      "loss": 0.7525,
+      "step": 10400
+    },
+    {
+      "epoch": 31.26,
+      "grad_norm": 18.495691299438477,
+      "learning_rate": 8.957957957957958e-06,
+      "loss": 0.7291,
+      "step": 10410
+    },
+    {
+      "epoch": 31.29,
+      "grad_norm": 11.233891487121582,
+      "learning_rate": 8.956956956956958e-06,
+      "loss": 0.7199,
+      "step": 10420
+    },
+    {
+      "epoch": 31.32,
+      "grad_norm": 11.309521675109863,
+      "learning_rate": 8.955955955955957e-06,
+      "loss": 0.7664,
+      "step": 10430
+    },
+    {
+      "epoch": 31.35,
+      "grad_norm": 13.5640230178833,
+      "learning_rate": 8.954954954954956e-06,
+      "loss": 0.7566,
+      "step": 10440
+    },
+    {
+      "epoch": 31.38,
+      "grad_norm": 17.966054916381836,
+      "learning_rate": 8.953953953953954e-06,
+      "loss": 0.7663,
+      "step": 10450
+    },
+    {
+      "epoch": 31.41,
+      "grad_norm": 15.01924991607666,
+      "learning_rate": 8.952952952952953e-06,
+      "loss": 0.7395,
+      "step": 10460
+    },
+    {
+      "epoch": 31.44,
+      "grad_norm": 15.482879638671875,
+      "learning_rate": 8.951951951951953e-06,
+      "loss": 0.6937,
+      "step": 10470
+    },
+    {
+      "epoch": 31.47,
+      "grad_norm": 14.140779495239258,
+      "learning_rate": 8.950950950950952e-06,
+      "loss": 0.7769,
+      "step": 10480
+    },
+    {
+      "epoch": 31.5,
+      "grad_norm": 15.519713401794434,
+      "learning_rate": 8.94994994994995e-06,
+      "loss": 0.7388,
+      "step": 10490
+    },
+    {
+      "epoch": 31.53,
+      "grad_norm": 13.309924125671387,
+      "learning_rate": 8.94894894894895e-06,
+      "loss": 0.7968,
+      "step": 10500
+    },
+    {
+      "epoch": 31.56,
+      "grad_norm": 17.812463760375977,
+      "learning_rate": 8.94794794794795e-06,
+      "loss": 0.7677,
+      "step": 10510
+    },
+    {
+      "epoch": 31.59,
+      "grad_norm": 19.066875457763672,
+      "learning_rate": 8.946946946946948e-06,
+      "loss": 0.7271,
+      "step": 10520
+    },
+    {
+      "epoch": 31.62,
+      "grad_norm": 18.764406204223633,
+      "learning_rate": 8.945945945945946e-06,
+      "loss": 0.7753,
+      "step": 10530
+    },
+    {
+      "epoch": 31.65,
+      "grad_norm": 20.98076629638672,
+      "learning_rate": 8.944944944944945e-06,
+      "loss": 0.7288,
+      "step": 10540
+    },
+    {
+      "epoch": 31.68,
+      "grad_norm": 19.42600440979004,
+      "learning_rate": 8.943943943943945e-06,
+      "loss": 0.7835,
+      "step": 10550
+    },
+    {
+      "epoch": 31.71,
+      "grad_norm": 15.125018119812012,
+      "learning_rate": 8.942942942942944e-06,
+      "loss": 0.7244,
+      "step": 10560
+    },
+    {
+      "epoch": 31.74,
+      "grad_norm": 13.277347564697266,
+      "learning_rate": 8.941941941941942e-06,
+      "loss": 0.8209,
+      "step": 10570
+    },
+    {
+      "epoch": 31.77,
+      "grad_norm": 12.954193115234375,
+      "learning_rate": 8.940940940940943e-06,
+      "loss": 0.7931,
+      "step": 10580
+    },
+    {
+      "epoch": 31.8,
+      "grad_norm": 19.218364715576172,
+      "learning_rate": 8.939939939939941e-06,
+      "loss": 0.8124,
+      "step": 10590
+    },
+    {
+      "epoch": 31.83,
+      "grad_norm": 19.637117385864258,
+      "learning_rate": 8.93893893893894e-06,
+      "loss": 0.6632,
+      "step": 10600
+    },
+    {
+      "epoch": 31.86,
+      "grad_norm": 16.105588912963867,
+      "learning_rate": 8.937937937937939e-06,
+      "loss": 0.762,
+      "step": 10610
+    },
+    {
+      "epoch": 31.89,
+      "grad_norm": 15.068008422851562,
+      "learning_rate": 8.936936936936937e-06,
+      "loss": 0.7938,
+      "step": 10620
+    },
+    {
+      "epoch": 31.92,
+      "grad_norm": 13.940908432006836,
+      "learning_rate": 8.935935935935937e-06,
+      "loss": 0.7368,
+      "step": 10630
+    },
+    {
+      "epoch": 31.95,
+      "grad_norm": 13.348282814025879,
+      "learning_rate": 8.934934934934936e-06,
+      "loss": 0.7477,
+      "step": 10640
+    },
+    {
+      "epoch": 31.98,
+      "grad_norm": 16.640409469604492,
+      "learning_rate": 8.933933933933935e-06,
+      "loss": 0.7179,
+      "step": 10650
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.8628,
+      "eval_loss": 0.45731332898139954,
+      "eval_runtime": 12.6917,
+      "eval_samples_per_second": 787.917,
+      "eval_steps_per_second": 3.152,
+      "step": 10656
+    },
+    {
+      "epoch": 32.01,
+      "grad_norm": 12.540587425231934,
+      "learning_rate": 8.932932932932933e-06,
+      "loss": 0.633,
+      "step": 10660
+    },
+    {
+      "epoch": 32.04,
+      "grad_norm": 14.436250686645508,
+      "learning_rate": 8.931931931931933e-06,
+      "loss": 0.7423,
+      "step": 10670
+    },
+    {
+      "epoch": 32.07,
+      "grad_norm": 12.762358665466309,
+      "learning_rate": 8.93093093093093e-06,
+      "loss": 0.7614,
+      "step": 10680
+    },
+    {
+      "epoch": 32.1,
+      "grad_norm": 12.973264694213867,
+      "learning_rate": 8.92992992992993e-06,
+      "loss": 0.7105,
+      "step": 10690
+    },
+    {
+      "epoch": 32.13,
+      "grad_norm": 18.508798599243164,
+      "learning_rate": 8.92892892892893e-06,
+      "loss": 0.74,
+      "step": 10700
+    },
+    {
+      "epoch": 32.16,
+      "grad_norm": 21.682180404663086,
+      "learning_rate": 8.927927927927928e-06,
+      "loss": 0.7725,
+      "step": 10710
+    },
+    {
+      "epoch": 32.19,
+      "grad_norm": 16.946609497070312,
+      "learning_rate": 8.926926926926928e-06,
+      "loss": 0.7464,
+      "step": 10720
+    },
+    {
+      "epoch": 32.22,
+      "grad_norm": 16.109943389892578,
+      "learning_rate": 8.925925925925927e-06,
+      "loss": 0.7526,
+      "step": 10730
+    },
+    {
+      "epoch": 32.25,
+      "grad_norm": 26.805973052978516,
+      "learning_rate": 8.924924924924925e-06,
+      "loss": 0.8015,
+      "step": 10740
+    },
+    {
+      "epoch": 32.28,
+      "grad_norm": 16.171995162963867,
+      "learning_rate": 8.923923923923926e-06,
+      "loss": 0.7825,
+      "step": 10750
+    },
+    {
+      "epoch": 32.31,
+      "grad_norm": 13.513046264648438,
+      "learning_rate": 8.922922922922924e-06,
+      "loss": 0.7439,
+      "step": 10760
+    },
+    {
+      "epoch": 32.34,
+      "grad_norm": 17.154273986816406,
+      "learning_rate": 8.921921921921923e-06,
+      "loss": 0.6719,
+      "step": 10770
+    },
+    {
+      "epoch": 32.37,
+      "grad_norm": 12.631355285644531,
+      "learning_rate": 8.920920920920921e-06,
+      "loss": 0.7323,
+      "step": 10780
+    },
+    {
+      "epoch": 32.4,
+      "grad_norm": 13.410599708557129,
+      "learning_rate": 8.91991991991992e-06,
+      "loss": 0.7047,
+      "step": 10790
+    },
+    {
+      "epoch": 32.43,
+      "grad_norm": 20.0744686126709,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.7817,
+      "step": 10800
+    },
+    {
+      "epoch": 32.46,
+      "grad_norm": 14.753862380981445,
+      "learning_rate": 8.917917917917919e-06,
+      "loss": 0.7503,
+      "step": 10810
+    },
+    {
+      "epoch": 32.49,
+      "grad_norm": 17.139963150024414,
+      "learning_rate": 8.916916916916917e-06,
+      "loss": 0.632,
+      "step": 10820
+    },
+    {
+      "epoch": 32.52,
+      "grad_norm": 22.070852279663086,
+      "learning_rate": 8.915915915915918e-06,
+      "loss": 0.759,
+      "step": 10830
+    },
+    {
+      "epoch": 32.55,
+      "grad_norm": 14.3706693649292,
+      "learning_rate": 8.914914914914916e-06,
+      "loss": 0.7906,
+      "step": 10840
+    },
+    {
+      "epoch": 32.58,
+      "grad_norm": 13.108540534973145,
+      "learning_rate": 8.913913913913915e-06,
+      "loss": 0.7442,
+      "step": 10850
+    },
+    {
+      "epoch": 32.61,
+      "grad_norm": 14.135384559631348,
+      "learning_rate": 8.912912912912914e-06,
+      "loss": 0.765,
+      "step": 10860
+    },
+    {
+      "epoch": 32.64,
+      "grad_norm": 16.804826736450195,
+      "learning_rate": 8.911911911911912e-06,
+      "loss": 0.7468,
+      "step": 10870
+    },
+    {
+      "epoch": 32.67,
+      "grad_norm": 21.809114456176758,
+      "learning_rate": 8.91091091091091e-06,
+      "loss": 0.7202,
+      "step": 10880
+    },
+    {
+      "epoch": 32.7,
+      "grad_norm": 13.352616310119629,
+      "learning_rate": 8.909909909909911e-06,
+      "loss": 0.7129,
+      "step": 10890
+    },
+    {
+      "epoch": 32.73,
+      "grad_norm": 18.08527946472168,
+      "learning_rate": 8.90890890890891e-06,
+      "loss": 0.7206,
+      "step": 10900
+    },
+    {
+      "epoch": 32.76,
+      "grad_norm": 13.541923522949219,
+      "learning_rate": 8.907907907907908e-06,
+      "loss": 0.7004,
+      "step": 10910
+    },
+    {
+      "epoch": 32.79,
+      "grad_norm": 18.027904510498047,
+      "learning_rate": 8.906906906906909e-06,
+      "loss": 0.7492,
+      "step": 10920
+    },
+    {
+      "epoch": 32.82,
+      "grad_norm": 14.436220169067383,
+      "learning_rate": 8.905905905905905e-06,
+      "loss": 0.7447,
+      "step": 10930
+    },
+    {
+      "epoch": 32.85,
+      "grad_norm": 16.82992172241211,
+      "learning_rate": 8.904904904904906e-06,
+      "loss": 0.7365,
+      "step": 10940
+    },
+    {
+      "epoch": 32.88,
+      "grad_norm": 15.015875816345215,
+      "learning_rate": 8.903903903903904e-06,
+      "loss": 0.7351,
+      "step": 10950
+    },
+    {
+      "epoch": 32.91,
+      "grad_norm": 14.806081771850586,
+      "learning_rate": 8.902902902902903e-06,
+      "loss": 0.7421,
+      "step": 10960
+    },
+    {
+      "epoch": 32.94,
+      "grad_norm": 21.481706619262695,
+      "learning_rate": 8.901901901901903e-06,
+      "loss": 0.7305,
+      "step": 10970
+    },
+    {
+      "epoch": 32.97,
+      "grad_norm": 19.971282958984375,
+      "learning_rate": 8.900900900900902e-06,
+      "loss": 0.6866,
+      "step": 10980
+    },
+    {
+      "epoch": 33.0,
+      "eval_accuracy": 0.8606,
+      "eval_loss": 0.45673778653144836,
+      "eval_runtime": 12.7087,
+      "eval_samples_per_second": 786.86,
+      "eval_steps_per_second": 3.147,
+      "step": 10989
+    },
+    {
+      "epoch": 33.0,
+      "grad_norm": 12.800230979919434,
+      "learning_rate": 8.8998998998999e-06,
+      "loss": 0.646,
+      "step": 10990
+    },
+    {
+      "epoch": 33.03,
+      "grad_norm": 17.8346004486084,
+      "learning_rate": 8.8988988988989e-06,
+      "loss": 0.8046,
+      "step": 11000
+    },
+    {
+      "epoch": 33.06,
+      "grad_norm": 18.317232131958008,
+      "learning_rate": 8.8978978978979e-06,
+      "loss": 0.7373,
+      "step": 11010
+    },
+    {
+      "epoch": 33.09,
+      "grad_norm": 15.877395629882812,
+      "learning_rate": 8.896896896896898e-06,
+      "loss": 0.6893,
+      "step": 11020
+    },
+    {
+      "epoch": 33.12,
+      "grad_norm": 14.688508033752441,
+      "learning_rate": 8.895895895895896e-06,
+      "loss": 0.7672,
+      "step": 11030
+    },
+    {
+      "epoch": 33.15,
+      "grad_norm": 13.723471641540527,
+      "learning_rate": 8.894894894894895e-06,
+      "loss": 0.7018,
+      "step": 11040
+    },
+    {
+      "epoch": 33.18,
+      "grad_norm": 24.826631546020508,
+      "learning_rate": 8.893893893893895e-06,
+      "loss": 0.7106,
+      "step": 11050
+    },
+    {
+      "epoch": 33.21,
+      "grad_norm": 21.92152976989746,
+      "learning_rate": 8.892892892892894e-06,
+      "loss": 0.8031,
+      "step": 11060
+    },
+    {
+      "epoch": 33.24,
+      "grad_norm": 15.498889923095703,
+      "learning_rate": 8.891891891891893e-06,
+      "loss": 0.7772,
+      "step": 11070
+    },
+    {
+      "epoch": 33.27,
+      "grad_norm": 17.40592384338379,
+      "learning_rate": 8.890890890890893e-06,
+      "loss": 0.6772,
+      "step": 11080
+    },
+    {
+      "epoch": 33.3,
+      "grad_norm": 12.944879531860352,
+      "learning_rate": 8.889889889889891e-06,
+      "loss": 0.6895,
+      "step": 11090
+    },
+    {
+      "epoch": 33.33,
+      "grad_norm": 18.65313148498535,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.751,
+      "step": 11100
+    },
+    {
+      "epoch": 33.36,
+      "grad_norm": 15.92672348022461,
+      "learning_rate": 8.887887887887889e-06,
+      "loss": 0.7871,
+      "step": 11110
+    },
+    {
+      "epoch": 33.39,
+      "grad_norm": 14.8170166015625,
+      "learning_rate": 8.886886886886887e-06,
+      "loss": 0.6701,
+      "step": 11120
+    },
+    {
+      "epoch": 33.42,
+      "grad_norm": 15.525805473327637,
+      "learning_rate": 8.885885885885886e-06,
+      "loss": 0.7314,
+      "step": 11130
+    },
+    {
+      "epoch": 33.45,
+      "grad_norm": 13.073753356933594,
+      "learning_rate": 8.884884884884886e-06,
+      "loss": 0.7163,
+      "step": 11140
+    },
+    {
+      "epoch": 33.48,
+      "grad_norm": 18.035778045654297,
+      "learning_rate": 8.883883883883885e-06,
+      "loss": 0.6964,
+      "step": 11150
+    },
+    {
+      "epoch": 33.51,
+      "grad_norm": 20.28458595275879,
+      "learning_rate": 8.882882882882883e-06,
+      "loss": 0.7479,
+      "step": 11160
+    },
+    {
+      "epoch": 33.54,
+      "grad_norm": 15.934008598327637,
+      "learning_rate": 8.881881881881884e-06,
+      "loss": 0.7319,
+      "step": 11170
+    },
+    {
+      "epoch": 33.57,
+      "grad_norm": 21.8447208404541,
+      "learning_rate": 8.88088088088088e-06,
+      "loss": 0.774,
+      "step": 11180
+    },
+    {
+      "epoch": 33.6,
+      "grad_norm": 10.164825439453125,
+      "learning_rate": 8.87987987987988e-06,
+      "loss": 0.7126,
+      "step": 11190
+    },
+    {
+      "epoch": 33.63,
+      "grad_norm": 14.525633811950684,
+      "learning_rate": 8.87887887887888e-06,
+      "loss": 0.6872,
+      "step": 11200
+    },
+    {
+      "epoch": 33.66,
+      "grad_norm": 22.03354835510254,
+      "learning_rate": 8.877877877877878e-06,
+      "loss": 0.7237,
+      "step": 11210
+    },
+    {
+      "epoch": 33.69,
+      "grad_norm": 13.250273704528809,
+      "learning_rate": 8.876876876876878e-06,
+      "loss": 0.8269,
+      "step": 11220
+    },
+    {
+      "epoch": 33.72,
+      "grad_norm": 33.590572357177734,
+      "learning_rate": 8.875875875875877e-06,
+      "loss": 0.7296,
+      "step": 11230
+    },
+    {
+      "epoch": 33.75,
+      "grad_norm": 15.577840805053711,
+      "learning_rate": 8.874874874874875e-06,
+      "loss": 0.7686,
+      "step": 11240
+    },
+    {
+      "epoch": 33.78,
+      "grad_norm": 19.518146514892578,
+      "learning_rate": 8.873873873873876e-06,
+      "loss": 0.8281,
+      "step": 11250
+    },
+    {
+      "epoch": 33.81,
+      "grad_norm": 15.385758399963379,
+      "learning_rate": 8.872872872872874e-06,
+      "loss": 0.7358,
+      "step": 11260
+    },
+    {
+      "epoch": 33.84,
+      "grad_norm": 18.950332641601562,
+      "learning_rate": 8.871871871871873e-06,
+      "loss": 0.7626,
+      "step": 11270
+    },
+    {
+      "epoch": 33.87,
+      "grad_norm": 18.48717498779297,
+      "learning_rate": 8.870870870870871e-06,
+      "loss": 0.7698,
+      "step": 11280
+    },
+    {
+      "epoch": 33.9,
+      "grad_norm": 17.826658248901367,
+      "learning_rate": 8.86986986986987e-06,
+      "loss": 0.7064,
+      "step": 11290
+    },
+    {
+      "epoch": 33.93,
+      "grad_norm": 13.667326927185059,
+      "learning_rate": 8.86886886886887e-06,
+      "loss": 0.7336,
+      "step": 11300
+    },
+    {
+      "epoch": 33.96,
+      "grad_norm": 14.063284873962402,
+      "learning_rate": 8.867867867867869e-06,
+      "loss": 0.7627,
+      "step": 11310
+    },
+    {
+      "epoch": 33.99,
+      "grad_norm": 10.707511901855469,
+      "learning_rate": 8.866866866866868e-06,
+      "loss": 0.7002,
+      "step": 11320
+    },
+    {
+      "epoch": 34.0,
+      "eval_accuracy": 0.8576,
+      "eval_loss": 0.4672236442565918,
+      "eval_runtime": 12.5384,
+      "eval_samples_per_second": 797.551,
+      "eval_steps_per_second": 3.19,
+      "step": 11322
+    },
+    {
+      "epoch": 34.02,
+      "grad_norm": 18.776241302490234,
+      "learning_rate": 8.865865865865866e-06,
+      "loss": 0.8133,
+      "step": 11330
+    },
+    {
+      "epoch": 34.05,
+      "grad_norm": 16.304729461669922,
+      "learning_rate": 8.864864864864866e-06,
+      "loss": 0.7111,
+      "step": 11340
+    },
+    {
+      "epoch": 34.08,
+      "grad_norm": 18.15487289428711,
+      "learning_rate": 8.863863863863863e-06,
+      "loss": 0.7658,
+      "step": 11350
+    },
+    {
+      "epoch": 34.11,
+      "grad_norm": 20.110403060913086,
+      "learning_rate": 8.862862862862864e-06,
+      "loss": 0.7248,
+      "step": 11360
+    },
+    {
+      "epoch": 34.14,
+      "grad_norm": 19.843303680419922,
+      "learning_rate": 8.861861861861862e-06,
+      "loss": 0.7505,
+      "step": 11370
+    },
+    {
+      "epoch": 34.17,
+      "grad_norm": 14.428262710571289,
+      "learning_rate": 8.86086086086086e-06,
+      "loss": 0.7215,
+      "step": 11380
+    },
+    {
+      "epoch": 34.2,
+      "grad_norm": 14.41454792022705,
+      "learning_rate": 8.859859859859861e-06,
+      "loss": 0.7939,
+      "step": 11390
+    },
+    {
+      "epoch": 34.23,
+      "grad_norm": 17.413043975830078,
+      "learning_rate": 8.85885885885886e-06,
+      "loss": 0.7546,
+      "step": 11400
+    },
+    {
+      "epoch": 34.26,
+      "grad_norm": 16.91671371459961,
+      "learning_rate": 8.857857857857858e-06,
+      "loss": 0.6689,
+      "step": 11410
+    },
+    {
+      "epoch": 34.29,
+      "grad_norm": 15.457772254943848,
+      "learning_rate": 8.856856856856859e-06,
+      "loss": 0.7038,
+      "step": 11420
+    },
+    {
+      "epoch": 34.32,
+      "grad_norm": 14.812891006469727,
+      "learning_rate": 8.855855855855855e-06,
+      "loss": 0.7306,
+      "step": 11430
+    },
+    {
+      "epoch": 34.35,
+      "grad_norm": 13.014900207519531,
+      "learning_rate": 8.854854854854856e-06,
+      "loss": 0.683,
+      "step": 11440
+    },
+    {
+      "epoch": 34.38,
+      "grad_norm": 15.84607982635498,
+      "learning_rate": 8.853853853853854e-06,
+      "loss": 0.7169,
+      "step": 11450
+    },
+    {
+      "epoch": 34.41,
+      "grad_norm": 17.296056747436523,
+      "learning_rate": 8.852852852852853e-06,
+      "loss": 0.6834,
+      "step": 11460
+    },
+    {
+      "epoch": 34.44,
+      "grad_norm": 19.145584106445312,
+      "learning_rate": 8.851851851851853e-06,
+      "loss": 0.7491,
+      "step": 11470
+    },
+    {
+      "epoch": 34.47,
+      "grad_norm": 12.988202095031738,
+      "learning_rate": 8.850850850850852e-06,
+      "loss": 0.7727,
+      "step": 11480
+    },
+    {
+      "epoch": 34.5,
+      "grad_norm": 15.30691909790039,
+      "learning_rate": 8.84984984984985e-06,
+      "loss": 0.6978,
+      "step": 11490
+    },
+    {
+      "epoch": 34.53,
+      "grad_norm": 20.13016700744629,
+      "learning_rate": 8.84884884884885e-06,
+      "loss": 0.6725,
+      "step": 11500
+    },
+    {
+      "epoch": 34.56,
+      "grad_norm": 14.221665382385254,
+      "learning_rate": 8.84784784784785e-06,
+      "loss": 0.7926,
+      "step": 11510
+    },
+    {
+      "epoch": 34.59,
+      "grad_norm": 20.517425537109375,
+      "learning_rate": 8.846846846846848e-06,
+      "loss": 0.7829,
+      "step": 11520
+    },
+    {
+      "epoch": 34.62,
+      "grad_norm": 16.56093978881836,
+      "learning_rate": 8.845845845845847e-06,
+      "loss": 0.7553,
+      "step": 11530
+    },
+    {
+      "epoch": 34.65,
+      "grad_norm": 15.996752738952637,
+      "learning_rate": 8.844844844844845e-06,
+      "loss": 0.6587,
+      "step": 11540
+    },
+    {
+      "epoch": 34.68,
+      "grad_norm": 19.360082626342773,
+      "learning_rate": 8.843843843843844e-06,
+      "loss": 0.7418,
+      "step": 11550
+    },
+    {
+      "epoch": 34.71,
+      "grad_norm": 14.33364486694336,
+      "learning_rate": 8.842842842842844e-06,
+      "loss": 0.7477,
+      "step": 11560
+    },
+    {
+      "epoch": 34.74,
+      "grad_norm": 18.078895568847656,
+      "learning_rate": 8.841841841841843e-06,
+      "loss": 0.7428,
+      "step": 11570
+    },
+    {
+      "epoch": 34.77,
+      "grad_norm": 22.46721649169922,
+      "learning_rate": 8.840840840840841e-06,
+      "loss": 0.7594,
+      "step": 11580
+    },
+    {
+      "epoch": 34.8,
+      "grad_norm": 19.704814910888672,
+      "learning_rate": 8.839839839839841e-06,
+      "loss": 0.7462,
+      "step": 11590
+    },
+    {
+      "epoch": 34.83,
+      "grad_norm": 16.586517333984375,
+      "learning_rate": 8.838838838838838e-06,
+      "loss": 0.7261,
+      "step": 11600
+    },
+    {
+      "epoch": 34.86,
+      "grad_norm": 14.008977890014648,
+      "learning_rate": 8.837837837837839e-06,
+      "loss": 0.7063,
+      "step": 11610
+    },
+    {
+      "epoch": 34.89,
+      "grad_norm": 23.218713760375977,
+      "learning_rate": 8.836836836836837e-06,
+      "loss": 0.7072,
+      "step": 11620
+    },
+    {
+      "epoch": 34.92,
+      "grad_norm": 12.442441940307617,
+      "learning_rate": 8.835835835835836e-06,
+      "loss": 0.7531,
+      "step": 11630
+    },
+    {
+      "epoch": 34.95,
+      "grad_norm": 14.057740211486816,
+      "learning_rate": 8.834834834834836e-06,
+      "loss": 0.7379,
+      "step": 11640
+    },
+    {
+      "epoch": 34.98,
+      "grad_norm": 15.848073959350586,
+      "learning_rate": 8.833833833833835e-06,
+      "loss": 0.7499,
+      "step": 11650
+    },
+    {
+      "epoch": 35.0,
+      "eval_accuracy": 0.8611,
+      "eval_loss": 0.46235302090644836,
+      "eval_runtime": 12.973,
+      "eval_samples_per_second": 770.832,
+      "eval_steps_per_second": 3.083,
+      "step": 11655
+    },
+    {
+      "epoch": 35.02,
+      "grad_norm": 15.122843742370605,
+      "learning_rate": 8.832832832832833e-06,
+      "loss": 0.6307,
+      "step": 11660
+    },
+    {
+      "epoch": 35.05,
+      "grad_norm": 16.84096336364746,
+      "learning_rate": 8.831831831831834e-06,
+      "loss": 0.6986,
+      "step": 11670
+    },
+    {
+      "epoch": 35.08,
+      "grad_norm": 11.740920066833496,
+      "learning_rate": 8.83083083083083e-06,
+      "loss": 0.7522,
+      "step": 11680
+    },
+    {
+      "epoch": 35.11,
+      "grad_norm": 13.660318374633789,
+      "learning_rate": 8.82982982982983e-06,
+      "loss": 0.7223,
+      "step": 11690
+    },
+    {
+      "epoch": 35.14,
+      "grad_norm": 18.145273208618164,
+      "learning_rate": 8.82882882882883e-06,
+      "loss": 0.7319,
+      "step": 11700
+    },
+    {
+      "epoch": 35.17,
+      "grad_norm": 17.636823654174805,
+      "learning_rate": 8.827827827827828e-06,
+      "loss": 0.7589,
+      "step": 11710
+    },
+    {
+      "epoch": 35.2,
+      "grad_norm": 11.691192626953125,
+      "learning_rate": 8.826826826826828e-06,
+      "loss": 0.7016,
+      "step": 11720
+    },
+    {
+      "epoch": 35.23,
+      "grad_norm": 18.20246696472168,
+      "learning_rate": 8.825825825825827e-06,
+      "loss": 0.7366,
+      "step": 11730
+    },
+    {
+      "epoch": 35.26,
+      "grad_norm": 13.31651782989502,
+      "learning_rate": 8.824824824824825e-06,
+      "loss": 0.766,
+      "step": 11740
+    },
+    {
+      "epoch": 35.29,
+      "grad_norm": 17.355192184448242,
+      "learning_rate": 8.823823823823826e-06,
+      "loss": 0.6873,
+      "step": 11750
+    },
+    {
+      "epoch": 35.32,
+      "grad_norm": 17.020751953125,
+      "learning_rate": 8.822822822822824e-06,
+      "loss": 0.7825,
+      "step": 11760
+    },
+    {
+      "epoch": 35.35,
+      "grad_norm": 20.819473266601562,
+      "learning_rate": 8.821821821821823e-06,
+      "loss": 0.7451,
+      "step": 11770
+    },
+    {
+      "epoch": 35.38,
+      "grad_norm": 15.498403549194336,
+      "learning_rate": 8.820820820820822e-06,
+      "loss": 0.7573,
+      "step": 11780
+    },
+    {
+      "epoch": 35.41,
+      "grad_norm": 23.820362091064453,
+      "learning_rate": 8.81981981981982e-06,
+      "loss": 0.803,
+      "step": 11790
+    },
+    {
+      "epoch": 35.44,
+      "grad_norm": 14.88689136505127,
+      "learning_rate": 8.818818818818819e-06,
+      "loss": 0.7374,
+      "step": 11800
+    },
+    {
+      "epoch": 35.47,
+      "grad_norm": 20.722328186035156,
+      "learning_rate": 8.817817817817819e-06,
+      "loss": 0.748,
+      "step": 11810
+    },
+    {
+      "epoch": 35.5,
+      "grad_norm": 17.91551971435547,
+      "learning_rate": 8.816816816816818e-06,
+      "loss": 0.7649,
+      "step": 11820
+    },
+    {
+      "epoch": 35.53,
+      "grad_norm": 16.54551124572754,
+      "learning_rate": 8.815815815815816e-06,
+      "loss": 0.7282,
+      "step": 11830
+    },
+    {
+      "epoch": 35.56,
+      "grad_norm": 17.026153564453125,
+      "learning_rate": 8.814814814814817e-06,
+      "loss": 0.7405,
+      "step": 11840
+    },
+    {
+      "epoch": 35.59,
+      "grad_norm": 15.0831880569458,
+      "learning_rate": 8.813813813813813e-06,
+      "loss": 0.7831,
+      "step": 11850
+    },
+    {
+      "epoch": 35.62,
+      "grad_norm": 12.430713653564453,
+      "learning_rate": 8.812812812812814e-06,
+      "loss": 0.6733,
+      "step": 11860
+    },
+    {
+      "epoch": 35.65,
+      "grad_norm": 11.541810035705566,
+      "learning_rate": 8.811811811811812e-06,
+      "loss": 0.7605,
+      "step": 11870
+    },
+    {
+      "epoch": 35.68,
+      "grad_norm": 18.88973617553711,
+      "learning_rate": 8.810810810810811e-06,
+      "loss": 0.7422,
+      "step": 11880
+    },
+    {
+      "epoch": 35.71,
+      "grad_norm": 17.730871200561523,
+      "learning_rate": 8.809809809809811e-06,
+      "loss": 0.7544,
+      "step": 11890
+    },
+    {
+      "epoch": 35.74,
+      "grad_norm": 18.9675235748291,
+      "learning_rate": 8.80880880880881e-06,
+      "loss": 0.7171,
+      "step": 11900
+    },
+    {
+      "epoch": 35.77,
+      "grad_norm": 13.948833465576172,
+      "learning_rate": 8.807807807807808e-06,
+      "loss": 0.6782,
+      "step": 11910
+    },
+    {
+      "epoch": 35.8,
+      "grad_norm": 18.81002426147461,
+      "learning_rate": 8.806806806806809e-06,
+      "loss": 0.7638,
+      "step": 11920
+    },
+    {
+      "epoch": 35.83,
+      "grad_norm": 15.860320091247559,
+      "learning_rate": 8.805805805805806e-06,
+      "loss": 0.7025,
+      "step": 11930
+    },
+    {
+      "epoch": 35.86,
+      "grad_norm": 17.77602195739746,
+      "learning_rate": 8.804804804804806e-06,
+      "loss": 0.7484,
+      "step": 11940
+    },
+    {
+      "epoch": 35.89,
+      "grad_norm": 25.042753219604492,
+      "learning_rate": 8.803803803803804e-06,
+      "loss": 0.7014,
+      "step": 11950
+    },
+    {
+      "epoch": 35.92,
+      "grad_norm": 13.907393455505371,
+      "learning_rate": 8.802802802802803e-06,
+      "loss": 0.7804,
+      "step": 11960
+    },
+    {
+      "epoch": 35.95,
+      "grad_norm": 13.40766429901123,
+      "learning_rate": 8.801801801801803e-06,
+      "loss": 0.6443,
+      "step": 11970
+    },
+    {
+      "epoch": 35.98,
+      "grad_norm": 11.657933235168457,
+      "learning_rate": 8.800800800800802e-06,
+      "loss": 0.7393,
+      "step": 11980
+    },
+    {
+      "epoch": 36.0,
+      "eval_accuracy": 0.8604,
+      "eval_loss": 0.45787227153778076,
+      "eval_runtime": 12.6531,
+      "eval_samples_per_second": 790.318,
+      "eval_steps_per_second": 3.161,
+      "step": 11988
+    },
+    {
+      "epoch": 36.01,
+      "grad_norm": 11.238343238830566,
+      "learning_rate": 8.7997997997998e-06,
+      "loss": 0.7178,
+      "step": 11990
+    },
+    {
+      "epoch": 36.04,
+      "grad_norm": 12.2100191116333,
+      "learning_rate": 8.798798798798799e-06,
+      "loss": 0.6748,
+      "step": 12000
+    },
+    {
+      "epoch": 36.07,
+      "grad_norm": 15.70413875579834,
+      "learning_rate": 8.797797797797798e-06,
+      "loss": 0.77,
+      "step": 12010
+    },
+    {
+      "epoch": 36.1,
+      "grad_norm": 13.942218780517578,
+      "learning_rate": 8.796796796796796e-06,
+      "loss": 0.7306,
+      "step": 12020
+    },
+    {
+      "epoch": 36.13,
+      "grad_norm": 15.168832778930664,
+      "learning_rate": 8.795795795795797e-06,
+      "loss": 0.746,
+      "step": 12030
+    },
+    {
+      "epoch": 36.16,
+      "grad_norm": 13.083436012268066,
+      "learning_rate": 8.794794794794795e-06,
+      "loss": 0.7263,
+      "step": 12040
+    },
+    {
+      "epoch": 36.19,
+      "grad_norm": 11.604787826538086,
+      "learning_rate": 8.793793793793794e-06,
+      "loss": 0.6678,
+      "step": 12050
+    },
+    {
+      "epoch": 36.22,
+      "grad_norm": 12.220256805419922,
+      "learning_rate": 8.792792792792794e-06,
+      "loss": 0.701,
+      "step": 12060
+    },
+    {
+      "epoch": 36.25,
+      "grad_norm": 15.767328262329102,
+      "learning_rate": 8.791791791791793e-06,
+      "loss": 0.6641,
+      "step": 12070
+    },
+    {
+      "epoch": 36.28,
+      "grad_norm": 18.16026496887207,
+      "learning_rate": 8.790790790790791e-06,
+      "loss": 0.6757,
+      "step": 12080
+    },
+    {
+      "epoch": 36.31,
+      "grad_norm": 15.411042213439941,
+      "learning_rate": 8.789789789789792e-06,
+      "loss": 0.6962,
+      "step": 12090
+    },
+    {
+      "epoch": 36.34,
+      "grad_norm": 16.6982421875,
+      "learning_rate": 8.788788788788788e-06,
+      "loss": 0.655,
+      "step": 12100
+    },
+    {
+      "epoch": 36.37,
+      "grad_norm": 17.151487350463867,
+      "learning_rate": 8.787787787787789e-06,
+      "loss": 0.7209,
+      "step": 12110
+    },
+    {
+      "epoch": 36.4,
+      "grad_norm": 11.250988960266113,
+      "learning_rate": 8.786786786786787e-06,
+      "loss": 0.7736,
+      "step": 12120
+    },
+    {
+      "epoch": 36.43,
+      "grad_norm": 15.72909927368164,
+      "learning_rate": 8.785785785785786e-06,
+      "loss": 0.727,
+      "step": 12130
+    },
+    {
+      "epoch": 36.46,
+      "grad_norm": 15.478221893310547,
+      "learning_rate": 8.784784784784786e-06,
+      "loss": 0.6861,
+      "step": 12140
+    },
+    {
+      "epoch": 36.49,
+      "grad_norm": 14.448071479797363,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 0.6771,
+      "step": 12150
+    },
+    {
+      "epoch": 36.52,
+      "grad_norm": 18.77178192138672,
+      "learning_rate": 8.782782782782783e-06,
+      "loss": 0.6514,
+      "step": 12160
+    },
+    {
+      "epoch": 36.55,
+      "grad_norm": 19.86296844482422,
+      "learning_rate": 8.781781781781784e-06,
+      "loss": 0.7216,
+      "step": 12170
+    },
+    {
+      "epoch": 36.58,
+      "grad_norm": 14.47397518157959,
+      "learning_rate": 8.78078078078078e-06,
+      "loss": 0.7528,
+      "step": 12180
+    },
+    {
+      "epoch": 36.61,
+      "grad_norm": 18.463701248168945,
+      "learning_rate": 8.779779779779781e-06,
+      "loss": 0.7306,
+      "step": 12190
+    },
+    {
+      "epoch": 36.64,
+      "grad_norm": 22.52227783203125,
+      "learning_rate": 8.77877877877878e-06,
+      "loss": 0.7225,
+      "step": 12200
+    },
+    {
+      "epoch": 36.67,
+      "grad_norm": 19.450387954711914,
+      "learning_rate": 8.777777777777778e-06,
+      "loss": 0.7652,
+      "step": 12210
+    },
+    {
+      "epoch": 36.7,
+      "grad_norm": 16.372949600219727,
+      "learning_rate": 8.776776776776778e-06,
+      "loss": 0.6711,
+      "step": 12220
+    },
+    {
+      "epoch": 36.73,
+      "grad_norm": 16.941741943359375,
+      "learning_rate": 8.775775775775777e-06,
+      "loss": 0.6673,
+      "step": 12230
+    },
+    {
+      "epoch": 36.76,
+      "grad_norm": 17.055347442626953,
+      "learning_rate": 8.774774774774776e-06,
+      "loss": 0.6923,
+      "step": 12240
+    },
+    {
+      "epoch": 36.79,
+      "grad_norm": 12.24468994140625,
+      "learning_rate": 8.773773773773774e-06,
+      "loss": 0.7552,
+      "step": 12250
+    },
+    {
+      "epoch": 36.82,
+      "grad_norm": 15.246345520019531,
+      "learning_rate": 8.772772772772773e-06,
+      "loss": 0.6756,
+      "step": 12260
+    },
+    {
+      "epoch": 36.85,
+      "grad_norm": 14.248589515686035,
+      "learning_rate": 8.771771771771771e-06,
+      "loss": 0.697,
+      "step": 12270
+    },
+    {
+      "epoch": 36.88,
+      "grad_norm": 18.042526245117188,
+      "learning_rate": 8.770770770770772e-06,
+      "loss": 0.7176,
+      "step": 12280
+    },
+    {
+      "epoch": 36.91,
+      "grad_norm": 17.553979873657227,
+      "learning_rate": 8.76976976976977e-06,
+      "loss": 0.7256,
+      "step": 12290
+    },
+    {
+      "epoch": 36.94,
+      "grad_norm": 14.7088623046875,
+      "learning_rate": 8.768768768768769e-06,
+      "loss": 0.7202,
+      "step": 12300
+    },
+    {
+      "epoch": 36.97,
+      "grad_norm": 13.636119842529297,
+      "learning_rate": 8.767767767767769e-06,
+      "loss": 0.6976,
+      "step": 12310
+    },
+    {
+      "epoch": 37.0,
+      "grad_norm": 13.1280517578125,
+      "learning_rate": 8.766766766766768e-06,
+      "loss": 0.7393,
+      "step": 12320
+    },
+    {
+      "epoch": 37.0,
+      "eval_accuracy": 0.8619,
+      "eval_loss": 0.4559956192970276,
+      "eval_runtime": 12.8598,
+      "eval_samples_per_second": 777.618,
+      "eval_steps_per_second": 3.11,
+      "step": 12321
+    },
+    {
+      "epoch": 37.03,
+      "grad_norm": 19.153818130493164,
+      "learning_rate": 8.765765765765766e-06,
+      "loss": 0.7586,
+      "step": 12330
+    },
+    {
+      "epoch": 37.06,
+      "grad_norm": 17.784103393554688,
+      "learning_rate": 8.764764764764767e-06,
+      "loss": 0.689,
+      "step": 12340
+    },
+    {
+      "epoch": 37.09,
+      "grad_norm": 20.15745735168457,
+      "learning_rate": 8.763763763763763e-06,
+      "loss": 0.7198,
+      "step": 12350
+    },
+    {
+      "epoch": 37.12,
+      "grad_norm": 19.974891662597656,
+      "learning_rate": 8.762762762762764e-06,
+      "loss": 0.7454,
+      "step": 12360
+    },
+    {
+      "epoch": 37.15,
+      "grad_norm": 15.03307819366455,
+      "learning_rate": 8.761761761761762e-06,
+      "loss": 0.6477,
+      "step": 12370
+    },
+    {
+      "epoch": 37.18,
+      "grad_norm": 18.829687118530273,
+      "learning_rate": 8.760760760760761e-06,
+      "loss": 0.7508,
+      "step": 12380
+    },
+    {
+      "epoch": 37.21,
+      "grad_norm": 24.866214752197266,
+      "learning_rate": 8.759759759759761e-06,
+      "loss": 0.7404,
+      "step": 12390
+    },
+    {
+      "epoch": 37.24,
+      "grad_norm": 12.424365043640137,
+      "learning_rate": 8.75875875875876e-06,
+      "loss": 0.6448,
+      "step": 12400
+    },
+    {
+      "epoch": 37.27,
+      "grad_norm": 17.853824615478516,
+      "learning_rate": 8.757757757757758e-06,
+      "loss": 0.6773,
+      "step": 12410
+    },
+    {
+      "epoch": 37.3,
+      "grad_norm": 15.64295768737793,
+      "learning_rate": 8.756756756756759e-06,
+      "loss": 0.6744,
+      "step": 12420
+    },
+    {
+      "epoch": 37.33,
+      "grad_norm": 13.909880638122559,
+      "learning_rate": 8.755755755755756e-06,
+      "loss": 0.7238,
+      "step": 12430
+    },
+    {
+      "epoch": 37.36,
+      "grad_norm": 12.074225425720215,
+      "learning_rate": 8.754754754754756e-06,
+      "loss": 0.7357,
+      "step": 12440
+    },
+    {
+      "epoch": 37.39,
+      "grad_norm": 20.037227630615234,
+      "learning_rate": 8.753753753753755e-06,
+      "loss": 0.6693,
+      "step": 12450
+    },
+    {
+      "epoch": 37.42,
+      "grad_norm": 14.504189491271973,
+      "learning_rate": 8.752752752752753e-06,
+      "loss": 0.6185,
+      "step": 12460
+    },
+    {
+      "epoch": 37.45,
+      "grad_norm": 14.980822563171387,
+      "learning_rate": 8.751751751751752e-06,
+      "loss": 0.7665,
+      "step": 12470
+    },
+    {
+      "epoch": 37.48,
+      "grad_norm": 14.017216682434082,
+      "learning_rate": 8.750750750750752e-06,
+      "loss": 0.7558,
+      "step": 12480
+    },
+    {
+      "epoch": 37.51,
+      "grad_norm": 18.15026092529297,
+      "learning_rate": 8.74974974974975e-06,
+      "loss": 0.7032,
+      "step": 12490
+    },
+    {
+      "epoch": 37.54,
+      "grad_norm": 21.63968849182129,
+      "learning_rate": 8.74874874874875e-06,
+      "loss": 0.7766,
+      "step": 12500
+    },
+    {
+      "epoch": 37.57,
+      "grad_norm": 13.396917343139648,
+      "learning_rate": 8.747747747747748e-06,
+      "loss": 0.7048,
+      "step": 12510
+    },
+    {
+      "epoch": 37.6,
+      "grad_norm": 12.383674621582031,
+      "learning_rate": 8.746746746746746e-06,
+      "loss": 0.7738,
+      "step": 12520
+    },
+    {
+      "epoch": 37.63,
+      "grad_norm": 14.813749313354492,
+      "learning_rate": 8.745745745745747e-06,
+      "loss": 0.7799,
+      "step": 12530
+    },
+    {
+      "epoch": 37.66,
+      "grad_norm": 18.46799659729004,
+      "learning_rate": 8.744744744744745e-06,
+      "loss": 0.7278,
+      "step": 12540
+    },
+    {
+      "epoch": 37.69,
+      "grad_norm": 16.44928550720215,
+      "learning_rate": 8.743743743743744e-06,
+      "loss": 0.6939,
+      "step": 12550
+    },
+    {
+      "epoch": 37.72,
+      "grad_norm": 16.80510902404785,
+      "learning_rate": 8.742742742742744e-06,
+      "loss": 0.6232,
+      "step": 12560
+    },
+    {
+      "epoch": 37.75,
+      "grad_norm": 17.623991012573242,
+      "learning_rate": 8.741741741741743e-06,
+      "loss": 0.6502,
+      "step": 12570
+    },
+    {
+      "epoch": 37.78,
+      "grad_norm": 22.835899353027344,
+      "learning_rate": 8.740740740740741e-06,
+      "loss": 0.6595,
+      "step": 12580
+    },
+    {
+      "epoch": 37.81,
+      "grad_norm": 17.97762680053711,
+      "learning_rate": 8.739739739739742e-06,
+      "loss": 0.6918,
+      "step": 12590
+    },
+    {
+      "epoch": 37.84,
+      "grad_norm": 16.164525985717773,
+      "learning_rate": 8.738738738738739e-06,
+      "loss": 0.7216,
+      "step": 12600
+    },
+    {
+      "epoch": 37.87,
+      "grad_norm": 19.715900421142578,
+      "learning_rate": 8.737737737737739e-06,
+      "loss": 0.727,
+      "step": 12610
+    },
+    {
+      "epoch": 37.9,
+      "grad_norm": 17.584728240966797,
+      "learning_rate": 8.736736736736737e-06,
+      "loss": 0.714,
+      "step": 12620
+    },
+    {
+      "epoch": 37.93,
+      "grad_norm": 14.52590274810791,
+      "learning_rate": 8.735735735735736e-06,
+      "loss": 0.7141,
+      "step": 12630
+    },
+    {
+      "epoch": 37.96,
+      "grad_norm": 11.749207496643066,
+      "learning_rate": 8.734734734734736e-06,
+      "loss": 0.7204,
+      "step": 12640
+    },
+    {
+      "epoch": 37.99,
+      "grad_norm": 12.749289512634277,
+      "learning_rate": 8.733733733733735e-06,
+      "loss": 0.7599,
+      "step": 12650
+    },
+    {
+      "epoch": 38.0,
+      "eval_accuracy": 0.8637,
+      "eval_loss": 0.4503398835659027,
+      "eval_runtime": 12.8021,
+      "eval_samples_per_second": 781.122,
+      "eval_steps_per_second": 3.124,
+      "step": 12654
+    },
+    {
+      "epoch": 38.02,
+      "grad_norm": 12.482909202575684,
+      "learning_rate": 8.732732732732733e-06,
+      "loss": 0.6813,
+      "step": 12660
+    },
+    {
+      "epoch": 38.05,
+      "grad_norm": 14.705399513244629,
+      "learning_rate": 8.731731731731734e-06,
+      "loss": 0.716,
+      "step": 12670
+    },
+    {
+      "epoch": 38.08,
+      "grad_norm": 12.497817039489746,
+      "learning_rate": 8.73073073073073e-06,
+      "loss": 0.6871,
+      "step": 12680
+    },
+    {
+      "epoch": 38.11,
+      "grad_norm": 18.887226104736328,
+      "learning_rate": 8.72972972972973e-06,
+      "loss": 0.6538,
+      "step": 12690
+    },
+    {
+      "epoch": 38.14,
+      "grad_norm": 26.343616485595703,
+      "learning_rate": 8.72872872872873e-06,
+      "loss": 0.6999,
+      "step": 12700
+    },
+    {
+      "epoch": 38.17,
+      "grad_norm": 17.42749786376953,
+      "learning_rate": 8.727727727727728e-06,
+      "loss": 0.6758,
+      "step": 12710
+    },
+    {
+      "epoch": 38.2,
+      "grad_norm": 16.925424575805664,
+      "learning_rate": 8.726726726726727e-06,
+      "loss": 0.7202,
+      "step": 12720
+    },
+    {
+      "epoch": 38.23,
+      "grad_norm": 19.516830444335938,
+      "learning_rate": 8.725725725725727e-06,
+      "loss": 0.6942,
+      "step": 12730
+    },
+    {
+      "epoch": 38.26,
+      "grad_norm": 23.34222984313965,
+      "learning_rate": 8.724724724724726e-06,
+      "loss": 0.7196,
+      "step": 12740
+    },
+    {
+      "epoch": 38.29,
+      "grad_norm": 14.984414100646973,
+      "learning_rate": 8.723723723723724e-06,
+      "loss": 0.7194,
+      "step": 12750
+    },
+    {
+      "epoch": 38.32,
+      "grad_norm": 17.17688751220703,
+      "learning_rate": 8.722722722722723e-06,
+      "loss": 0.7218,
+      "step": 12760
+    },
+    {
+      "epoch": 38.35,
+      "grad_norm": 13.632336616516113,
+      "learning_rate": 8.721721721721721e-06,
+      "loss": 0.686,
+      "step": 12770
+    },
+    {
+      "epoch": 38.38,
+      "grad_norm": 12.847562789916992,
+      "learning_rate": 8.720720720720722e-06,
+      "loss": 0.6567,
+      "step": 12780
+    },
+    {
+      "epoch": 38.41,
+      "grad_norm": 17.647014617919922,
+      "learning_rate": 8.71971971971972e-06,
+      "loss": 0.673,
+      "step": 12790
+    },
+    {
+      "epoch": 38.44,
+      "grad_norm": 14.931132316589355,
+      "learning_rate": 8.718718718718719e-06,
+      "loss": 0.668,
+      "step": 12800
+    },
+    {
+      "epoch": 38.47,
+      "grad_norm": 17.570600509643555,
+      "learning_rate": 8.71771771771772e-06,
+      "loss": 0.7211,
+      "step": 12810
+    },
+    {
+      "epoch": 38.5,
+      "grad_norm": 19.52607536315918,
+      "learning_rate": 8.716716716716718e-06,
+      "loss": 0.7171,
+      "step": 12820
+    },
+    {
+      "epoch": 38.53,
+      "grad_norm": 17.393821716308594,
+      "learning_rate": 8.715715715715716e-06,
+      "loss": 0.656,
+      "step": 12830
+    },
+    {
+      "epoch": 38.56,
+      "grad_norm": 17.661428451538086,
+      "learning_rate": 8.714714714714717e-06,
+      "loss": 0.6668,
+      "step": 12840
+    },
+    {
+      "epoch": 38.59,
+      "grad_norm": 19.984769821166992,
+      "learning_rate": 8.713713713713714e-06,
+      "loss": 0.6828,
+      "step": 12850
+    },
+    {
+      "epoch": 38.62,
+      "grad_norm": 17.77810287475586,
+      "learning_rate": 8.712712712712714e-06,
+      "loss": 0.7671,
+      "step": 12860
+    },
+    {
+      "epoch": 38.65,
+      "grad_norm": 14.88648509979248,
+      "learning_rate": 8.711711711711712e-06,
+      "loss": 0.6844,
+      "step": 12870
+    },
+    {
+      "epoch": 38.68,
+      "grad_norm": 13.530986785888672,
+      "learning_rate": 8.710710710710711e-06,
+      "loss": 0.7069,
+      "step": 12880
+    },
+    {
+      "epoch": 38.71,
+      "grad_norm": 16.827951431274414,
+      "learning_rate": 8.709709709709711e-06,
+      "loss": 0.7527,
+      "step": 12890
+    },
+    {
+      "epoch": 38.74,
+      "grad_norm": 18.00787353515625,
+      "learning_rate": 8.70870870870871e-06,
+      "loss": 0.6821,
+      "step": 12900
+    },
+    {
+      "epoch": 38.77,
+      "grad_norm": 11.284110069274902,
+      "learning_rate": 8.707707707707708e-06,
+      "loss": 0.642,
+      "step": 12910
+    },
+    {
+      "epoch": 38.8,
+      "grad_norm": 14.642751693725586,
+      "learning_rate": 8.706706706706707e-06,
+      "loss": 0.7072,
+      "step": 12920
+    },
+    {
+      "epoch": 38.83,
+      "grad_norm": 13.521146774291992,
+      "learning_rate": 8.705705705705706e-06,
+      "loss": 0.6929,
+      "step": 12930
+    },
+    {
+      "epoch": 38.86,
+      "grad_norm": 15.691649436950684,
+      "learning_rate": 8.704704704704704e-06,
+      "loss": 0.6839,
+      "step": 12940
+    },
+    {
+      "epoch": 38.89,
+      "grad_norm": 15.077134132385254,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 0.7288,
+      "step": 12950
+    },
+    {
+      "epoch": 38.92,
+      "grad_norm": 18.418928146362305,
+      "learning_rate": 8.702702702702703e-06,
+      "loss": 0.7192,
+      "step": 12960
+    },
+    {
+      "epoch": 38.95,
+      "grad_norm": 16.348697662353516,
+      "learning_rate": 8.701701701701702e-06,
+      "loss": 0.7042,
+      "step": 12970
+    },
+    {
+      "epoch": 38.98,
+      "grad_norm": 13.3355073928833,
+      "learning_rate": 8.700700700700702e-06,
+      "loss": 0.6636,
+      "step": 12980
+    },
+    {
+      "epoch": 39.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.4542168974876404,
+      "eval_runtime": 12.7582,
+      "eval_samples_per_second": 783.808,
+      "eval_steps_per_second": 3.135,
+      "step": 12987
+    },
+    {
+      "epoch": 39.01,
+      "grad_norm": 17.700868606567383,
+      "learning_rate": 8.6996996996997e-06,
+      "loss": 0.6941,
+      "step": 12990
+    },
+    {
+      "epoch": 39.04,
+      "grad_norm": 22.615550994873047,
+      "learning_rate": 8.6986986986987e-06,
+      "loss": 0.6983,
+      "step": 13000
+    },
+    {
+      "epoch": 39.07,
+      "grad_norm": 19.436965942382812,
+      "learning_rate": 8.697697697697698e-06,
+      "loss": 0.6828,
+      "step": 13010
+    },
+    {
+      "epoch": 39.1,
+      "grad_norm": 12.391670227050781,
+      "learning_rate": 8.696696696696696e-06,
+      "loss": 0.7454,
+      "step": 13020
+    },
+    {
+      "epoch": 39.13,
+      "grad_norm": 14.891891479492188,
+      "learning_rate": 8.695695695695697e-06,
+      "loss": 0.6627,
+      "step": 13030
+    },
+    {
+      "epoch": 39.16,
+      "grad_norm": 13.117813110351562,
+      "learning_rate": 8.694694694694695e-06,
+      "loss": 0.6927,
+      "step": 13040
+    },
+    {
+      "epoch": 39.19,
+      "grad_norm": 15.762950897216797,
+      "learning_rate": 8.693693693693694e-06,
+      "loss": 0.6707,
+      "step": 13050
+    },
+    {
+      "epoch": 39.22,
+      "grad_norm": 16.58308219909668,
+      "learning_rate": 8.692692692692694e-06,
+      "loss": 0.6483,
+      "step": 13060
+    },
+    {
+      "epoch": 39.25,
+      "grad_norm": 19.617237091064453,
+      "learning_rate": 8.691691691691693e-06,
+      "loss": 0.6582,
+      "step": 13070
+    },
+    {
+      "epoch": 39.28,
+      "grad_norm": 13.417759895324707,
+      "learning_rate": 8.690690690690691e-06,
+      "loss": 0.6729,
+      "step": 13080
+    },
+    {
+      "epoch": 39.31,
+      "grad_norm": 13.217031478881836,
+      "learning_rate": 8.689689689689692e-06,
+      "loss": 0.7673,
+      "step": 13090
+    },
+    {
+      "epoch": 39.34,
+      "grad_norm": 18.13391876220703,
+      "learning_rate": 8.688688688688689e-06,
+      "loss": 0.7142,
+      "step": 13100
+    },
+    {
+      "epoch": 39.37,
+      "grad_norm": 15.779520988464355,
+      "learning_rate": 8.687687687687689e-06,
+      "loss": 0.7541,
+      "step": 13110
+    },
+    {
+      "epoch": 39.4,
+      "grad_norm": 14.55923080444336,
+      "learning_rate": 8.686686686686687e-06,
+      "loss": 0.7059,
+      "step": 13120
+    },
+    {
+      "epoch": 39.43,
+      "grad_norm": 15.300395965576172,
+      "learning_rate": 8.685685685685686e-06,
+      "loss": 0.7028,
+      "step": 13130
+    },
+    {
+      "epoch": 39.46,
+      "grad_norm": 12.668709754943848,
+      "learning_rate": 8.684684684684686e-06,
+      "loss": 0.7071,
+      "step": 13140
+    },
+    {
+      "epoch": 39.49,
+      "grad_norm": 27.27690887451172,
+      "learning_rate": 8.683683683683685e-06,
+      "loss": 0.7162,
+      "step": 13150
+    },
+    {
+      "epoch": 39.52,
+      "grad_norm": 24.240915298461914,
+      "learning_rate": 8.682682682682684e-06,
+      "loss": 0.6125,
+      "step": 13160
+    },
+    {
+      "epoch": 39.55,
+      "grad_norm": 13.969755172729492,
+      "learning_rate": 8.681681681681682e-06,
+      "loss": 0.7409,
+      "step": 13170
+    },
+    {
+      "epoch": 39.58,
+      "grad_norm": 17.5896053314209,
+      "learning_rate": 8.68068068068068e-06,
+      "loss": 0.755,
+      "step": 13180
+    },
+    {
+      "epoch": 39.61,
+      "grad_norm": 15.106500625610352,
+      "learning_rate": 8.67967967967968e-06,
+      "loss": 0.7074,
+      "step": 13190
+    },
+    {
+      "epoch": 39.64,
+      "grad_norm": 13.774001121520996,
+      "learning_rate": 8.67867867867868e-06,
+      "loss": 0.7052,
+      "step": 13200
+    },
+    {
+      "epoch": 39.67,
+      "grad_norm": 17.810007095336914,
+      "learning_rate": 8.677677677677678e-06,
+      "loss": 0.692,
+      "step": 13210
+    },
+    {
+      "epoch": 39.7,
+      "grad_norm": 12.47335147857666,
+      "learning_rate": 8.676676676676677e-06,
+      "loss": 0.7536,
+      "step": 13220
+    },
+    {
+      "epoch": 39.73,
+      "grad_norm": 17.86687660217285,
+      "learning_rate": 8.675675675675677e-06,
+      "loss": 0.7456,
+      "step": 13230
+    },
+    {
+      "epoch": 39.76,
+      "grad_norm": 13.085111618041992,
+      "learning_rate": 8.674674674674676e-06,
+      "loss": 0.7243,
+      "step": 13240
+    },
+    {
+      "epoch": 39.79,
+      "grad_norm": 15.906510353088379,
+      "learning_rate": 8.673673673673674e-06,
+      "loss": 0.7421,
+      "step": 13250
+    },
+    {
+      "epoch": 39.82,
+      "grad_norm": 17.313884735107422,
+      "learning_rate": 8.672672672672673e-06,
+      "loss": 0.7011,
+      "step": 13260
+    },
+    {
+      "epoch": 39.85,
+      "grad_norm": 21.422687530517578,
+      "learning_rate": 8.671671671671671e-06,
+      "loss": 0.7689,
+      "step": 13270
+    },
+    {
+      "epoch": 39.88,
+      "grad_norm": 13.853211402893066,
+      "learning_rate": 8.670670670670672e-06,
+      "loss": 0.6439,
+      "step": 13280
+    },
+    {
+      "epoch": 39.91,
+      "grad_norm": 14.36677074432373,
+      "learning_rate": 8.66966966966967e-06,
+      "loss": 0.6398,
+      "step": 13290
+    },
+    {
+      "epoch": 39.94,
+      "grad_norm": 14.41776180267334,
+      "learning_rate": 8.668668668668669e-06,
+      "loss": 0.6711,
+      "step": 13300
+    },
+    {
+      "epoch": 39.97,
+      "grad_norm": 15.397546768188477,
+      "learning_rate": 8.66766766766767e-06,
+      "loss": 0.7268,
+      "step": 13310
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 26.0655460357666,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.6759,
+      "step": 13320
+    },
+    {
+      "epoch": 40.0,
+      "eval_accuracy": 0.8631,
+      "eval_loss": 0.4483490586280823,
+      "eval_runtime": 12.7377,
+      "eval_samples_per_second": 785.072,
+      "eval_steps_per_second": 3.14,
+      "step": 13320
+    },
+    {
+      "epoch": 40.03,
+      "grad_norm": 16.34037208557129,
+      "learning_rate": 8.665665665665666e-06,
+      "loss": 0.7081,
+      "step": 13330
+    },
+    {
+      "epoch": 40.06,
+      "grad_norm": 22.965871810913086,
+      "learning_rate": 8.664664664664665e-06,
+      "loss": 0.7028,
+      "step": 13340
+    },
+    {
+      "epoch": 40.09,
+      "grad_norm": 15.480595588684082,
+      "learning_rate": 8.663663663663664e-06,
+      "loss": 0.6835,
+      "step": 13350
+    },
+    {
+      "epoch": 40.12,
+      "grad_norm": 20.957406997680664,
+      "learning_rate": 8.662662662662664e-06,
+      "loss": 0.7048,
+      "step": 13360
+    },
+    {
+      "epoch": 40.15,
+      "grad_norm": 17.77429962158203,
+      "learning_rate": 8.661661661661662e-06,
+      "loss": 0.6815,
+      "step": 13370
+    },
+    {
+      "epoch": 40.18,
+      "grad_norm": 15.408767700195312,
+      "learning_rate": 8.660660660660661e-06,
+      "loss": 0.7629,
+      "step": 13380
+    },
+    {
+      "epoch": 40.21,
+      "grad_norm": 29.371509552001953,
+      "learning_rate": 8.65965965965966e-06,
+      "loss": 0.6937,
+      "step": 13390
+    },
+    {
+      "epoch": 40.24,
+      "grad_norm": 18.90596580505371,
+      "learning_rate": 8.65865865865866e-06,
+      "loss": 0.7482,
+      "step": 13400
+    },
+    {
+      "epoch": 40.27,
+      "grad_norm": 15.754220008850098,
+      "learning_rate": 8.657657657657659e-06,
+      "loss": 0.7235,
+      "step": 13410
+    },
+    {
+      "epoch": 40.3,
+      "grad_norm": 14.71186637878418,
+      "learning_rate": 8.656656656656657e-06,
+      "loss": 0.641,
+      "step": 13420
+    },
+    {
+      "epoch": 40.33,
+      "grad_norm": 13.372477531433105,
+      "learning_rate": 8.655655655655656e-06,
+      "loss": 0.7081,
+      "step": 13430
+    },
+    {
+      "epoch": 40.36,
+      "grad_norm": 14.14633846282959,
+      "learning_rate": 8.654654654654654e-06,
+      "loss": 0.6236,
+      "step": 13440
+    },
+    {
+      "epoch": 40.39,
+      "grad_norm": 19.37156105041504,
+      "learning_rate": 8.653653653653655e-06,
+      "loss": 0.7461,
+      "step": 13450
+    },
+    {
+      "epoch": 40.42,
+      "grad_norm": 10.999704360961914,
+      "learning_rate": 8.652652652652653e-06,
+      "loss": 0.7108,
+      "step": 13460
+    },
+    {
+      "epoch": 40.45,
+      "grad_norm": 13.646711349487305,
+      "learning_rate": 8.651651651651652e-06,
+      "loss": 0.6774,
+      "step": 13470
+    },
+    {
+      "epoch": 40.48,
+      "grad_norm": 16.37213897705078,
+      "learning_rate": 8.650650650650652e-06,
+      "loss": 0.7388,
+      "step": 13480
+    },
+    {
+      "epoch": 40.51,
+      "grad_norm": 13.935249328613281,
+      "learning_rate": 8.64964964964965e-06,
+      "loss": 0.6764,
+      "step": 13490
+    },
+    {
+      "epoch": 40.54,
+      "grad_norm": 18.431129455566406,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.6639,
+      "step": 13500
+    },
+    {
+      "epoch": 40.57,
+      "grad_norm": 17.029029846191406,
+      "learning_rate": 8.647647647647648e-06,
+      "loss": 0.7311,
+      "step": 13510
+    },
+    {
+      "epoch": 40.6,
+      "grad_norm": 17.871564865112305,
+      "learning_rate": 8.646646646646646e-06,
+      "loss": 0.6722,
+      "step": 13520
+    },
+    {
+      "epoch": 40.63,
+      "grad_norm": 21.63091278076172,
+      "learning_rate": 8.645645645645647e-06,
+      "loss": 0.6511,
+      "step": 13530
+    },
+    {
+      "epoch": 40.66,
+      "grad_norm": 16.98485565185547,
+      "learning_rate": 8.644644644644645e-06,
+      "loss": 0.6225,
+      "step": 13540
+    },
+    {
+      "epoch": 40.69,
+      "grad_norm": 21.113874435424805,
+      "learning_rate": 8.643643643643644e-06,
+      "loss": 0.6879,
+      "step": 13550
+    },
+    {
+      "epoch": 40.72,
+      "grad_norm": 14.520386695861816,
+      "learning_rate": 8.642642642642644e-06,
+      "loss": 0.681,
+      "step": 13560
+    },
+    {
+      "epoch": 40.75,
+      "grad_norm": 18.337249755859375,
+      "learning_rate": 8.641641641641643e-06,
+      "loss": 0.6918,
+      "step": 13570
+    },
+    {
+      "epoch": 40.78,
+      "grad_norm": 17.143888473510742,
+      "learning_rate": 8.640640640640641e-06,
+      "loss": 0.6627,
+      "step": 13580
+    },
+    {
+      "epoch": 40.81,
+      "grad_norm": 16.210668563842773,
+      "learning_rate": 8.63963963963964e-06,
+      "loss": 0.6589,
+      "step": 13590
+    },
+    {
+      "epoch": 40.84,
+      "grad_norm": 12.302651405334473,
+      "learning_rate": 8.638638638638639e-06,
+      "loss": 0.7005,
+      "step": 13600
+    },
+    {
+      "epoch": 40.87,
+      "grad_norm": 15.272514343261719,
+      "learning_rate": 8.637637637637637e-06,
+      "loss": 0.6784,
+      "step": 13610
+    },
+    {
+      "epoch": 40.9,
+      "grad_norm": 16.107824325561523,
+      "learning_rate": 8.636636636636638e-06,
+      "loss": 0.6779,
+      "step": 13620
+    },
+    {
+      "epoch": 40.93,
+      "grad_norm": 10.67597770690918,
+      "learning_rate": 8.635635635635636e-06,
+      "loss": 0.6982,
+      "step": 13630
+    },
+    {
+      "epoch": 40.96,
+      "grad_norm": 16.74396514892578,
+      "learning_rate": 8.634634634634635e-06,
+      "loss": 0.7181,
+      "step": 13640
+    },
+    {
+      "epoch": 40.99,
+      "grad_norm": 15.14286994934082,
+      "learning_rate": 8.633633633633635e-06,
+      "loss": 0.7266,
+      "step": 13650
+    },
+    {
+      "epoch": 41.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.44839540123939514,
+      "eval_runtime": 12.8276,
+      "eval_samples_per_second": 779.568,
+      "eval_steps_per_second": 3.118,
+      "step": 13653
+    },
+    {
+      "epoch": 41.02,
+      "grad_norm": 11.654528617858887,
+      "learning_rate": 8.632632632632634e-06,
+      "loss": 0.6292,
+      "step": 13660
+    },
+    {
+      "epoch": 41.05,
+      "grad_norm": 11.669693946838379,
+      "learning_rate": 8.631631631631632e-06,
+      "loss": 0.6264,
+      "step": 13670
+    },
+    {
+      "epoch": 41.08,
+      "grad_norm": 24.02885627746582,
+      "learning_rate": 8.63063063063063e-06,
+      "loss": 0.6173,
+      "step": 13680
+    },
+    {
+      "epoch": 41.11,
+      "grad_norm": 13.989689826965332,
+      "learning_rate": 8.62962962962963e-06,
+      "loss": 0.6185,
+      "step": 13690
+    },
+    {
+      "epoch": 41.14,
+      "grad_norm": 14.894391059875488,
+      "learning_rate": 8.62862862862863e-06,
+      "loss": 0.6867,
+      "step": 13700
+    },
+    {
+      "epoch": 41.17,
+      "grad_norm": 12.18368911743164,
+      "learning_rate": 8.627627627627628e-06,
+      "loss": 0.7027,
+      "step": 13710
+    },
+    {
+      "epoch": 41.2,
+      "grad_norm": 10.826661109924316,
+      "learning_rate": 8.626626626626627e-06,
+      "loss": 0.5933,
+      "step": 13720
+    },
+    {
+      "epoch": 41.23,
+      "grad_norm": 15.728558540344238,
+      "learning_rate": 8.625625625625627e-06,
+      "loss": 0.6609,
+      "step": 13730
+    },
+    {
+      "epoch": 41.26,
+      "grad_norm": 12.73317813873291,
+      "learning_rate": 8.624624624624626e-06,
+      "loss": 0.6357,
+      "step": 13740
+    },
+    {
+      "epoch": 41.29,
+      "grad_norm": 20.110111236572266,
+      "learning_rate": 8.623623623623624e-06,
+      "loss": 0.6726,
+      "step": 13750
+    },
+    {
+      "epoch": 41.32,
+      "grad_norm": 11.981986999511719,
+      "learning_rate": 8.622622622622623e-06,
+      "loss": 0.6711,
+      "step": 13760
+    },
+    {
+      "epoch": 41.35,
+      "grad_norm": 18.302762985229492,
+      "learning_rate": 8.621621621621622e-06,
+      "loss": 0.6916,
+      "step": 13770
+    },
+    {
+      "epoch": 41.38,
+      "grad_norm": 14.573165893554688,
+      "learning_rate": 8.620620620620622e-06,
+      "loss": 0.7553,
+      "step": 13780
+    },
+    {
+      "epoch": 41.41,
+      "grad_norm": 11.373902320861816,
+      "learning_rate": 8.61961961961962e-06,
+      "loss": 0.6533,
+      "step": 13790
+    },
+    {
+      "epoch": 41.44,
+      "grad_norm": 19.43169593811035,
+      "learning_rate": 8.618618618618619e-06,
+      "loss": 0.6569,
+      "step": 13800
+    },
+    {
+      "epoch": 41.47,
+      "grad_norm": 18.107942581176758,
+      "learning_rate": 8.61761761761762e-06,
+      "loss": 0.6466,
+      "step": 13810
+    },
+    {
+      "epoch": 41.5,
+      "grad_norm": 15.347994804382324,
+      "learning_rate": 8.616616616616618e-06,
+      "loss": 0.6784,
+      "step": 13820
+    },
+    {
+      "epoch": 41.53,
+      "grad_norm": 15.557960510253906,
+      "learning_rate": 8.615615615615616e-06,
+      "loss": 0.6951,
+      "step": 13830
+    },
+    {
+      "epoch": 41.56,
+      "grad_norm": 13.998001098632812,
+      "learning_rate": 8.614614614614615e-06,
+      "loss": 0.7094,
+      "step": 13840
+    },
+    {
+      "epoch": 41.59,
+      "grad_norm": 21.74806785583496,
+      "learning_rate": 8.613613613613614e-06,
+      "loss": 0.6718,
+      "step": 13850
+    },
+    {
+      "epoch": 41.62,
+      "grad_norm": 17.380090713500977,
+      "learning_rate": 8.612612612612612e-06,
+      "loss": 0.6722,
+      "step": 13860
+    },
+    {
+      "epoch": 41.65,
+      "grad_norm": 16.73099708557129,
+      "learning_rate": 8.611611611611613e-06,
+      "loss": 0.6847,
+      "step": 13870
+    },
+    {
+      "epoch": 41.68,
+      "grad_norm": 14.040267944335938,
+      "learning_rate": 8.610610610610611e-06,
+      "loss": 0.6534,
+      "step": 13880
+    },
+    {
+      "epoch": 41.71,
+      "grad_norm": 17.12603187561035,
+      "learning_rate": 8.60960960960961e-06,
+      "loss": 0.636,
+      "step": 13890
+    },
+    {
+      "epoch": 41.74,
+      "grad_norm": 13.38772964477539,
+      "learning_rate": 8.60860860860861e-06,
+      "loss": 0.7212,
+      "step": 13900
+    },
+    {
+      "epoch": 41.77,
+      "grad_norm": 13.89432430267334,
+      "learning_rate": 8.607607607607609e-06,
+      "loss": 0.7177,
+      "step": 13910
+    },
+    {
+      "epoch": 41.8,
+      "grad_norm": 16.896560668945312,
+      "learning_rate": 8.606606606606607e-06,
+      "loss": 0.7,
+      "step": 13920
+    },
+    {
+      "epoch": 41.83,
+      "grad_norm": 15.703804969787598,
+      "learning_rate": 8.605605605605606e-06,
+      "loss": 0.68,
+      "step": 13930
+    },
+    {
+      "epoch": 41.86,
+      "grad_norm": 19.20931053161621,
+      "learning_rate": 8.604604604604604e-06,
+      "loss": 0.6466,
+      "step": 13940
+    },
+    {
+      "epoch": 41.89,
+      "grad_norm": 15.076231002807617,
+      "learning_rate": 8.603603603603605e-06,
+      "loss": 0.6595,
+      "step": 13950
+    },
+    {
+      "epoch": 41.92,
+      "grad_norm": 19.755826950073242,
+      "learning_rate": 8.602602602602603e-06,
+      "loss": 0.6285,
+      "step": 13960
+    },
+    {
+      "epoch": 41.95,
+      "grad_norm": 16.200489044189453,
+      "learning_rate": 8.601601601601602e-06,
+      "loss": 0.7014,
+      "step": 13970
+    },
+    {
+      "epoch": 41.98,
+      "grad_norm": 13.048054695129395,
+      "learning_rate": 8.600600600600602e-06,
+      "loss": 0.6819,
+      "step": 13980
+    },
+    {
+      "epoch": 42.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.44531553983688354,
+      "eval_runtime": 12.8015,
+      "eval_samples_per_second": 781.157,
+      "eval_steps_per_second": 3.125,
+      "step": 13986
+    },
+    {
+      "epoch": 42.01,
+      "grad_norm": 18.29779624938965,
+      "learning_rate": 8.5995995995996e-06,
+      "loss": 0.6019,
+      "step": 13990
+    },
+    {
+      "epoch": 42.04,
+      "grad_norm": 16.439821243286133,
+      "learning_rate": 8.5985985985986e-06,
+      "loss": 0.6732,
+      "step": 14000
+    },
+    {
+      "epoch": 42.07,
+      "grad_norm": 16.085302352905273,
+      "learning_rate": 8.597597597597598e-06,
+      "loss": 0.724,
+      "step": 14010
+    },
+    {
+      "epoch": 42.1,
+      "grad_norm": 14.533585548400879,
+      "learning_rate": 8.596596596596597e-06,
+      "loss": 0.645,
+      "step": 14020
+    },
+    {
+      "epoch": 42.13,
+      "grad_norm": 21.371252059936523,
+      "learning_rate": 8.595595595595597e-06,
+      "loss": 0.6839,
+      "step": 14030
+    },
+    {
+      "epoch": 42.16,
+      "grad_norm": 20.06671905517578,
+      "learning_rate": 8.594594594594595e-06,
+      "loss": 0.6694,
+      "step": 14040
+    },
+    {
+      "epoch": 42.19,
+      "grad_norm": 12.474686622619629,
+      "learning_rate": 8.593593593593594e-06,
+      "loss": 0.7014,
+      "step": 14050
+    },
+    {
+      "epoch": 42.22,
+      "grad_norm": 16.817657470703125,
+      "learning_rate": 8.592592592592593e-06,
+      "loss": 0.6726,
+      "step": 14060
+    },
+    {
+      "epoch": 42.25,
+      "grad_norm": 15.75899600982666,
+      "learning_rate": 8.591591591591593e-06,
+      "loss": 0.7043,
+      "step": 14070
+    },
+    {
+      "epoch": 42.28,
+      "grad_norm": 17.445838928222656,
+      "learning_rate": 8.590590590590592e-06,
+      "loss": 0.7285,
+      "step": 14080
+    },
+    {
+      "epoch": 42.31,
+      "grad_norm": 15.50268840789795,
+      "learning_rate": 8.58958958958959e-06,
+      "loss": 0.6661,
+      "step": 14090
+    },
+    {
+      "epoch": 42.34,
+      "grad_norm": 15.050628662109375,
+      "learning_rate": 8.588588588588589e-06,
+      "loss": 0.7308,
+      "step": 14100
+    },
+    {
+      "epoch": 42.37,
+      "grad_norm": 17.504926681518555,
+      "learning_rate": 8.587587587587587e-06,
+      "loss": 0.6646,
+      "step": 14110
+    },
+    {
+      "epoch": 42.4,
+      "grad_norm": 16.50434684753418,
+      "learning_rate": 8.586586586586588e-06,
+      "loss": 0.7288,
+      "step": 14120
+    },
+    {
+      "epoch": 42.43,
+      "grad_norm": 17.286767959594727,
+      "learning_rate": 8.585585585585586e-06,
+      "loss": 0.6343,
+      "step": 14130
+    },
+    {
+      "epoch": 42.46,
+      "grad_norm": 18.94367218017578,
+      "learning_rate": 8.584584584584585e-06,
+      "loss": 0.653,
+      "step": 14140
+    },
+    {
+      "epoch": 42.49,
+      "grad_norm": 18.6459903717041,
+      "learning_rate": 8.583583583583585e-06,
+      "loss": 0.6394,
+      "step": 14150
+    },
+    {
+      "epoch": 42.52,
+      "grad_norm": 15.302431106567383,
+      "learning_rate": 8.582582582582584e-06,
+      "loss": 0.7375,
+      "step": 14160
+    },
+    {
+      "epoch": 42.55,
+      "grad_norm": 15.144591331481934,
+      "learning_rate": 8.581581581581582e-06,
+      "loss": 0.7067,
+      "step": 14170
+    },
+    {
+      "epoch": 42.58,
+      "grad_norm": 17.692134857177734,
+      "learning_rate": 8.580580580580581e-06,
+      "loss": 0.7066,
+      "step": 14180
+    },
+    {
+      "epoch": 42.61,
+      "grad_norm": 17.765499114990234,
+      "learning_rate": 8.57957957957958e-06,
+      "loss": 0.6299,
+      "step": 14190
+    },
+    {
+      "epoch": 42.64,
+      "grad_norm": 11.533652305603027,
+      "learning_rate": 8.57857857857858e-06,
+      "loss": 0.6854,
+      "step": 14200
+    },
+    {
+      "epoch": 42.67,
+      "grad_norm": 13.993724822998047,
+      "learning_rate": 8.577577577577578e-06,
+      "loss": 0.7785,
+      "step": 14210
+    },
+    {
+      "epoch": 42.7,
+      "grad_norm": 17.138479232788086,
+      "learning_rate": 8.576576576576577e-06,
+      "loss": 0.6751,
+      "step": 14220
+    },
+    {
+      "epoch": 42.73,
+      "grad_norm": 18.59140968322754,
+      "learning_rate": 8.575575575575577e-06,
+      "loss": 0.6469,
+      "step": 14230
+    },
+    {
+      "epoch": 42.76,
+      "grad_norm": 17.046478271484375,
+      "learning_rate": 8.574574574574576e-06,
+      "loss": 0.6916,
+      "step": 14240
+    },
+    {
+      "epoch": 42.79,
+      "grad_norm": 21.127248764038086,
+      "learning_rate": 8.573573573573574e-06,
+      "loss": 0.7079,
+      "step": 14250
+    },
+    {
+      "epoch": 42.82,
+      "grad_norm": 12.968728065490723,
+      "learning_rate": 8.572572572572573e-06,
+      "loss": 0.6931,
+      "step": 14260
+    },
+    {
+      "epoch": 42.85,
+      "grad_norm": 12.719165802001953,
+      "learning_rate": 8.571571571571572e-06,
+      "loss": 0.6553,
+      "step": 14270
+    },
+    {
+      "epoch": 42.88,
+      "grad_norm": 16.34621810913086,
+      "learning_rate": 8.570570570570572e-06,
+      "loss": 0.6198,
+      "step": 14280
+    },
+    {
+      "epoch": 42.91,
+      "grad_norm": 19.220603942871094,
+      "learning_rate": 8.56956956956957e-06,
+      "loss": 0.7594,
+      "step": 14290
+    },
+    {
+      "epoch": 42.94,
+      "grad_norm": 17.396108627319336,
+      "learning_rate": 8.568568568568569e-06,
+      "loss": 0.6612,
+      "step": 14300
+    },
+    {
+      "epoch": 42.97,
+      "grad_norm": 17.92281723022461,
+      "learning_rate": 8.567567567567568e-06,
+      "loss": 0.5912,
+      "step": 14310
+    },
+    {
+      "epoch": 43.0,
+      "eval_accuracy": 0.864,
+      "eval_loss": 0.44930997490882874,
+      "eval_runtime": 12.9537,
+      "eval_samples_per_second": 771.983,
+      "eval_steps_per_second": 3.088,
+      "step": 14319
+    },
+    {
+      "epoch": 43.0,
+      "grad_norm": 20.381256103515625,
+      "learning_rate": 8.566566566566568e-06,
+      "loss": 0.7875,
+      "step": 14320
+    },
+    {
+      "epoch": 43.03,
+      "grad_norm": 15.10037899017334,
+      "learning_rate": 8.565565565565567e-06,
+      "loss": 0.6471,
+      "step": 14330
+    },
+    {
+      "epoch": 43.06,
+      "grad_norm": 23.53544044494629,
+      "learning_rate": 8.564564564564565e-06,
+      "loss": 0.7124,
+      "step": 14340
+    },
+    {
+      "epoch": 43.09,
+      "grad_norm": 12.924009323120117,
+      "learning_rate": 8.563563563563564e-06,
+      "loss": 0.7007,
+      "step": 14350
+    },
+    {
+      "epoch": 43.12,
+      "grad_norm": 14.097970008850098,
+      "learning_rate": 8.562562562562562e-06,
+      "loss": 0.6386,
+      "step": 14360
+    },
+    {
+      "epoch": 43.15,
+      "grad_norm": 13.898308753967285,
+      "learning_rate": 8.561561561561563e-06,
+      "loss": 0.7065,
+      "step": 14370
+    },
+    {
+      "epoch": 43.18,
+      "grad_norm": 22.49960708618164,
+      "learning_rate": 8.560560560560561e-06,
+      "loss": 0.6298,
+      "step": 14380
+    },
+    {
+      "epoch": 43.21,
+      "grad_norm": 14.752483367919922,
+      "learning_rate": 8.55955955955956e-06,
+      "loss": 0.6568,
+      "step": 14390
+    },
+    {
+      "epoch": 43.24,
+      "grad_norm": 16.16034698486328,
+      "learning_rate": 8.55855855855856e-06,
+      "loss": 0.6402,
+      "step": 14400
+    },
+    {
+      "epoch": 43.27,
+      "grad_norm": 32.22293472290039,
+      "learning_rate": 8.557557557557559e-06,
+      "loss": 0.7523,
+      "step": 14410
+    },
+    {
+      "epoch": 43.3,
+      "grad_norm": 12.088333129882812,
+      "learning_rate": 8.556556556556557e-06,
+      "loss": 0.5934,
+      "step": 14420
+    },
+    {
+      "epoch": 43.33,
+      "grad_norm": 21.806533813476562,
+      "learning_rate": 8.555555555555556e-06,
+      "loss": 0.6815,
+      "step": 14430
+    },
+    {
+      "epoch": 43.36,
+      "grad_norm": 15.63686466217041,
+      "learning_rate": 8.554554554554554e-06,
+      "loss": 0.66,
+      "step": 14440
+    },
+    {
+      "epoch": 43.39,
+      "grad_norm": 16.214784622192383,
+      "learning_rate": 8.553553553553555e-06,
+      "loss": 0.7082,
+      "step": 14450
+    },
+    {
+      "epoch": 43.42,
+      "grad_norm": 15.74312973022461,
+      "learning_rate": 8.552552552552553e-06,
+      "loss": 0.685,
+      "step": 14460
+    },
+    {
+      "epoch": 43.45,
+      "grad_norm": 16.908809661865234,
+      "learning_rate": 8.551551551551552e-06,
+      "loss": 0.6638,
+      "step": 14470
+    },
+    {
+      "epoch": 43.48,
+      "grad_norm": 14.65440559387207,
+      "learning_rate": 8.550550550550552e-06,
+      "loss": 0.6949,
+      "step": 14480
+    },
+    {
+      "epoch": 43.51,
+      "grad_norm": 17.863666534423828,
+      "learning_rate": 8.549549549549551e-06,
+      "loss": 0.6373,
+      "step": 14490
+    },
+    {
+      "epoch": 43.54,
+      "grad_norm": 16.440378189086914,
+      "learning_rate": 8.54854854854855e-06,
+      "loss": 0.6862,
+      "step": 14500
+    },
+    {
+      "epoch": 43.57,
+      "grad_norm": 20.176227569580078,
+      "learning_rate": 8.547547547547548e-06,
+      "loss": 0.6744,
+      "step": 14510
+    },
+    {
+      "epoch": 43.6,
+      "grad_norm": 18.182384490966797,
+      "learning_rate": 8.546546546546547e-06,
+      "loss": 0.6158,
+      "step": 14520
+    },
+    {
+      "epoch": 43.63,
+      "grad_norm": 13.061956405639648,
+      "learning_rate": 8.545545545545545e-06,
+      "loss": 0.6501,
+      "step": 14530
+    },
+    {
+      "epoch": 43.66,
+      "grad_norm": 14.27139663696289,
+      "learning_rate": 8.544544544544546e-06,
+      "loss": 0.6133,
+      "step": 14540
+    },
+    {
+      "epoch": 43.69,
+      "grad_norm": 15.125067710876465,
+      "learning_rate": 8.543543543543544e-06,
+      "loss": 0.686,
+      "step": 14550
+    },
+    {
+      "epoch": 43.72,
+      "grad_norm": 22.890474319458008,
+      "learning_rate": 8.542542542542543e-06,
+      "loss": 0.5918,
+      "step": 14560
+    },
+    {
+      "epoch": 43.75,
+      "grad_norm": 15.257393836975098,
+      "learning_rate": 8.541541541541543e-06,
+      "loss": 0.7166,
+      "step": 14570
+    },
+    {
+      "epoch": 43.78,
+      "grad_norm": 16.940078735351562,
+      "learning_rate": 8.540540540540542e-06,
+      "loss": 0.6683,
+      "step": 14580
+    },
+    {
+      "epoch": 43.81,
+      "grad_norm": 16.10072135925293,
+      "learning_rate": 8.53953953953954e-06,
+      "loss": 0.6698,
+      "step": 14590
+    },
+    {
+      "epoch": 43.84,
+      "grad_norm": 15.821409225463867,
+      "learning_rate": 8.538538538538539e-06,
+      "loss": 0.7221,
+      "step": 14600
+    },
+    {
+      "epoch": 43.87,
+      "grad_norm": 17.96417808532715,
+      "learning_rate": 8.537537537537537e-06,
+      "loss": 0.677,
+      "step": 14610
+    },
+    {
+      "epoch": 43.9,
+      "grad_norm": 15.759181022644043,
+      "learning_rate": 8.536536536536538e-06,
+      "loss": 0.6919,
+      "step": 14620
+    },
+    {
+      "epoch": 43.93,
+      "grad_norm": 16.6920166015625,
+      "learning_rate": 8.535535535535536e-06,
+      "loss": 0.6148,
+      "step": 14630
+    },
+    {
+      "epoch": 43.96,
+      "grad_norm": 17.795061111450195,
+      "learning_rate": 8.534534534534535e-06,
+      "loss": 0.6573,
+      "step": 14640
+    },
+    {
+      "epoch": 43.99,
+      "grad_norm": 14.420255661010742,
+      "learning_rate": 8.533533533533535e-06,
+      "loss": 0.6803,
+      "step": 14650
+    },
+    {
+      "epoch": 44.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.44525375962257385,
+      "eval_runtime": 13.0886,
+      "eval_samples_per_second": 764.025,
+      "eval_steps_per_second": 3.056,
+      "step": 14652
+    },
+    {
+      "epoch": 44.02,
+      "grad_norm": 15.55947494506836,
+      "learning_rate": 8.532532532532534e-06,
+      "loss": 0.5994,
+      "step": 14660
+    },
+    {
+      "epoch": 44.05,
+      "grad_norm": 17.376937866210938,
+      "learning_rate": 8.531531531531532e-06,
+      "loss": 0.6174,
+      "step": 14670
+    },
+    {
+      "epoch": 44.08,
+      "grad_norm": 12.870214462280273,
+      "learning_rate": 8.530530530530531e-06,
+      "loss": 0.6422,
+      "step": 14680
+    },
+    {
+      "epoch": 44.11,
+      "grad_norm": 13.59985065460205,
+      "learning_rate": 8.52952952952953e-06,
+      "loss": 0.6884,
+      "step": 14690
+    },
+    {
+      "epoch": 44.14,
+      "grad_norm": 15.210004806518555,
+      "learning_rate": 8.52852852852853e-06,
+      "loss": 0.6767,
+      "step": 14700
+    },
+    {
+      "epoch": 44.17,
+      "grad_norm": 13.726428031921387,
+      "learning_rate": 8.527527527527528e-06,
+      "loss": 0.6895,
+      "step": 14710
+    },
+    {
+      "epoch": 44.2,
+      "grad_norm": 13.653573036193848,
+      "learning_rate": 8.526526526526527e-06,
+      "loss": 0.638,
+      "step": 14720
+    },
+    {
+      "epoch": 44.23,
+      "grad_norm": 13.787467002868652,
+      "learning_rate": 8.525525525525527e-06,
+      "loss": 0.6513,
+      "step": 14730
+    },
+    {
+      "epoch": 44.26,
+      "grad_norm": 15.06092643737793,
+      "learning_rate": 8.524524524524526e-06,
+      "loss": 0.667,
+      "step": 14740
+    },
+    {
+      "epoch": 44.29,
+      "grad_norm": 11.093276977539062,
+      "learning_rate": 8.523523523523524e-06,
+      "loss": 0.6214,
+      "step": 14750
+    },
+    {
+      "epoch": 44.32,
+      "grad_norm": 13.810016632080078,
+      "learning_rate": 8.522522522522523e-06,
+      "loss": 0.6762,
+      "step": 14760
+    },
+    {
+      "epoch": 44.35,
+      "grad_norm": 15.720680236816406,
+      "learning_rate": 8.521521521521522e-06,
+      "loss": 0.6101,
+      "step": 14770
+    },
+    {
+      "epoch": 44.38,
+      "grad_norm": 16.587196350097656,
+      "learning_rate": 8.52052052052052e-06,
+      "loss": 0.6182,
+      "step": 14780
+    },
+    {
+      "epoch": 44.41,
+      "grad_norm": 13.699054718017578,
+      "learning_rate": 8.51951951951952e-06,
+      "loss": 0.6334,
+      "step": 14790
+    },
+    {
+      "epoch": 44.44,
+      "grad_norm": 12.843461036682129,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.723,
+      "step": 14800
+    },
+    {
+      "epoch": 44.47,
+      "grad_norm": 17.749874114990234,
+      "learning_rate": 8.517517517517518e-06,
+      "loss": 0.6909,
+      "step": 14810
+    },
+    {
+      "epoch": 44.5,
+      "grad_norm": 18.269649505615234,
+      "learning_rate": 8.516516516516518e-06,
+      "loss": 0.6154,
+      "step": 14820
+    },
+    {
+      "epoch": 44.53,
+      "grad_norm": 14.030564308166504,
+      "learning_rate": 8.515515515515517e-06,
+      "loss": 0.6779,
+      "step": 14830
+    },
+    {
+      "epoch": 44.56,
+      "grad_norm": 14.909416198730469,
+      "learning_rate": 8.514514514514515e-06,
+      "loss": 0.6351,
+      "step": 14840
+    },
+    {
+      "epoch": 44.59,
+      "grad_norm": 24.80760383605957,
+      "learning_rate": 8.513513513513514e-06,
+      "loss": 0.6858,
+      "step": 14850
+    },
+    {
+      "epoch": 44.62,
+      "grad_norm": 12.409850120544434,
+      "learning_rate": 8.512512512512512e-06,
+      "loss": 0.593,
+      "step": 14860
+    },
+    {
+      "epoch": 44.65,
+      "grad_norm": 15.177338600158691,
+      "learning_rate": 8.511511511511513e-06,
+      "loss": 0.5928,
+      "step": 14870
+    },
+    {
+      "epoch": 44.68,
+      "grad_norm": 17.484256744384766,
+      "learning_rate": 8.510510510510511e-06,
+      "loss": 0.675,
+      "step": 14880
+    },
+    {
+      "epoch": 44.71,
+      "grad_norm": 14.007301330566406,
+      "learning_rate": 8.50950950950951e-06,
+      "loss": 0.644,
+      "step": 14890
+    },
+    {
+      "epoch": 44.74,
+      "grad_norm": 21.29224395751953,
+      "learning_rate": 8.50850850850851e-06,
+      "loss": 0.6996,
+      "step": 14900
+    },
+    {
+      "epoch": 44.77,
+      "grad_norm": 18.76585578918457,
+      "learning_rate": 8.507507507507509e-06,
+      "loss": 0.6909,
+      "step": 14910
+    },
+    {
+      "epoch": 44.8,
+      "grad_norm": 16.887725830078125,
+      "learning_rate": 8.506506506506507e-06,
+      "loss": 0.6274,
+      "step": 14920
+    },
+    {
+      "epoch": 44.83,
+      "grad_norm": 16.269027709960938,
+      "learning_rate": 8.505505505505506e-06,
+      "loss": 0.6791,
+      "step": 14930
+    },
+    {
+      "epoch": 44.86,
+      "grad_norm": 11.667265892028809,
+      "learning_rate": 8.504504504504505e-06,
+      "loss": 0.692,
+      "step": 14940
+    },
+    {
+      "epoch": 44.89,
+      "grad_norm": 12.998801231384277,
+      "learning_rate": 8.503503503503505e-06,
+      "loss": 0.6251,
+      "step": 14950
+    },
+    {
+      "epoch": 44.92,
+      "grad_norm": 20.578771591186523,
+      "learning_rate": 8.502502502502503e-06,
+      "loss": 0.6662,
+      "step": 14960
+    },
+    {
+      "epoch": 44.95,
+      "grad_norm": 13.840142250061035,
+      "learning_rate": 8.501501501501502e-06,
+      "loss": 0.7111,
+      "step": 14970
+    },
+    {
+      "epoch": 44.98,
+      "grad_norm": 14.712730407714844,
+      "learning_rate": 8.5005005005005e-06,
+      "loss": 0.6898,
+      "step": 14980
+    },
+    {
+      "epoch": 45.0,
+      "eval_accuracy": 0.8628,
+      "eval_loss": 0.44582709670066833,
+      "eval_runtime": 12.3919,
+      "eval_samples_per_second": 806.978,
+      "eval_steps_per_second": 3.228,
+      "step": 14985
+    },
+    {
+      "epoch": 45.02,
+      "grad_norm": 15.17032241821289,
+      "learning_rate": 8.499499499499501e-06,
+      "loss": 0.795,
+      "step": 14990
+    },
+    {
+      "epoch": 45.05,
+      "grad_norm": 14.758530616760254,
+      "learning_rate": 8.4984984984985e-06,
+      "loss": 0.6719,
+      "step": 15000
+    },
+    {
+      "epoch": 45.08,
+      "grad_norm": 14.033868789672852,
+      "learning_rate": 8.497497497497498e-06,
+      "loss": 0.6007,
+      "step": 15010
+    },
+    {
+      "epoch": 45.11,
+      "grad_norm": 20.025487899780273,
+      "learning_rate": 8.496496496496497e-06,
+      "loss": 0.6915,
+      "step": 15020
+    },
+    {
+      "epoch": 45.14,
+      "grad_norm": 18.14263153076172,
+      "learning_rate": 8.495495495495495e-06,
+      "loss": 0.6224,
+      "step": 15030
+    },
+    {
+      "epoch": 45.17,
+      "grad_norm": 18.517101287841797,
+      "learning_rate": 8.494494494494496e-06,
+      "loss": 0.6209,
+      "step": 15040
+    },
+    {
+      "epoch": 45.2,
+      "grad_norm": 11.368416786193848,
+      "learning_rate": 8.493493493493494e-06,
+      "loss": 0.6446,
+      "step": 15050
+    },
+    {
+      "epoch": 45.23,
+      "grad_norm": 14.640449523925781,
+      "learning_rate": 8.492492492492493e-06,
+      "loss": 0.6376,
+      "step": 15060
+    },
+    {
+      "epoch": 45.26,
+      "grad_norm": 20.73483657836914,
+      "learning_rate": 8.491491491491493e-06,
+      "loss": 0.6679,
+      "step": 15070
+    },
+    {
+      "epoch": 45.29,
+      "grad_norm": 15.753366470336914,
+      "learning_rate": 8.490490490490492e-06,
+      "loss": 0.6786,
+      "step": 15080
+    },
+    {
+      "epoch": 45.32,
+      "grad_norm": 12.244842529296875,
+      "learning_rate": 8.48948948948949e-06,
+      "loss": 0.6502,
+      "step": 15090
+    },
+    {
+      "epoch": 45.35,
+      "grad_norm": 16.75644874572754,
+      "learning_rate": 8.488488488488489e-06,
+      "loss": 0.6085,
+      "step": 15100
+    },
+    {
+      "epoch": 45.38,
+      "grad_norm": 12.139744758605957,
+      "learning_rate": 8.487487487487487e-06,
+      "loss": 0.7269,
+      "step": 15110
+    },
+    {
+      "epoch": 45.41,
+      "grad_norm": 13.40955924987793,
+      "learning_rate": 8.486486486486488e-06,
+      "loss": 0.6944,
+      "step": 15120
+    },
+    {
+      "epoch": 45.44,
+      "grad_norm": 22.504512786865234,
+      "learning_rate": 8.485485485485486e-06,
+      "loss": 0.6239,
+      "step": 15130
+    },
+    {
+      "epoch": 45.47,
+      "grad_norm": 14.809057235717773,
+      "learning_rate": 8.484484484484485e-06,
+      "loss": 0.6635,
+      "step": 15140
+    },
+    {
+      "epoch": 45.5,
+      "grad_norm": 20.3834285736084,
+      "learning_rate": 8.483483483483485e-06,
+      "loss": 0.672,
+      "step": 15150
+    },
+    {
+      "epoch": 45.53,
+      "grad_norm": 13.717203140258789,
+      "learning_rate": 8.482482482482484e-06,
+      "loss": 0.6436,
+      "step": 15160
+    },
+    {
+      "epoch": 45.56,
+      "grad_norm": 13.105717658996582,
+      "learning_rate": 8.481481481481482e-06,
+      "loss": 0.6458,
+      "step": 15170
+    },
+    {
+      "epoch": 45.59,
+      "grad_norm": 16.13928985595703,
+      "learning_rate": 8.480480480480481e-06,
+      "loss": 0.5647,
+      "step": 15180
+    },
+    {
+      "epoch": 45.62,
+      "grad_norm": 14.582718849182129,
+      "learning_rate": 8.47947947947948e-06,
+      "loss": 0.6806,
+      "step": 15190
+    },
+    {
+      "epoch": 45.65,
+      "grad_norm": 12.979229927062988,
+      "learning_rate": 8.47847847847848e-06,
+      "loss": 0.5787,
+      "step": 15200
+    },
+    {
+      "epoch": 45.68,
+      "grad_norm": 14.528862953186035,
+      "learning_rate": 8.477477477477478e-06,
+      "loss": 0.679,
+      "step": 15210
+    },
+    {
+      "epoch": 45.71,
+      "grad_norm": 14.974054336547852,
+      "learning_rate": 8.476476476476477e-06,
+      "loss": 0.6121,
+      "step": 15220
+    },
+    {
+      "epoch": 45.74,
+      "grad_norm": 19.096973419189453,
+      "learning_rate": 8.475475475475476e-06,
+      "loss": 0.5976,
+      "step": 15230
+    },
+    {
+      "epoch": 45.77,
+      "grad_norm": 12.103023529052734,
+      "learning_rate": 8.474474474474476e-06,
+      "loss": 0.6534,
+      "step": 15240
+    },
+    {
+      "epoch": 45.8,
+      "grad_norm": 15.813861846923828,
+      "learning_rate": 8.473473473473475e-06,
+      "loss": 0.652,
+      "step": 15250
+    },
+    {
+      "epoch": 45.83,
+      "grad_norm": 18.952356338500977,
+      "learning_rate": 8.472472472472473e-06,
+      "loss": 0.636,
+      "step": 15260
+    },
+    {
+      "epoch": 45.86,
+      "grad_norm": 21.465044021606445,
+      "learning_rate": 8.471471471471472e-06,
+      "loss": 0.619,
+      "step": 15270
+    },
+    {
+      "epoch": 45.89,
+      "grad_norm": 16.716754913330078,
+      "learning_rate": 8.47047047047047e-06,
+      "loss": 0.6996,
+      "step": 15280
+    },
+    {
+      "epoch": 45.92,
+      "grad_norm": 18.515457153320312,
+      "learning_rate": 8.46946946946947e-06,
+      "loss": 0.6817,
+      "step": 15290
+    },
+    {
+      "epoch": 45.95,
+      "grad_norm": 13.04875373840332,
+      "learning_rate": 8.46846846846847e-06,
+      "loss": 0.6051,
+      "step": 15300
+    },
+    {
+      "epoch": 45.98,
+      "grad_norm": 12.385298728942871,
+      "learning_rate": 8.467467467467468e-06,
+      "loss": 0.6312,
+      "step": 15310
+    },
+    {
+      "epoch": 46.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.44993892312049866,
+      "eval_runtime": 12.9287,
+      "eval_samples_per_second": 773.476,
+      "eval_steps_per_second": 3.094,
+      "step": 15318
+    },
+    {
+      "epoch": 46.01,
+      "grad_norm": 17.510120391845703,
+      "learning_rate": 8.466466466466468e-06,
+      "loss": 0.6119,
+      "step": 15320
+    },
+    {
+      "epoch": 46.04,
+      "grad_norm": 16.67719268798828,
+      "learning_rate": 8.465465465465467e-06,
+      "loss": 0.5838,
+      "step": 15330
+    },
+    {
+      "epoch": 46.07,
+      "grad_norm": 17.144563674926758,
+      "learning_rate": 8.464464464464465e-06,
+      "loss": 0.5972,
+      "step": 15340
+    },
+    {
+      "epoch": 46.1,
+      "grad_norm": 14.985879898071289,
+      "learning_rate": 8.463463463463464e-06,
+      "loss": 0.6636,
+      "step": 15350
+    },
+    {
+      "epoch": 46.13,
+      "grad_norm": 12.297521591186523,
+      "learning_rate": 8.462462462462462e-06,
+      "loss": 0.6748,
+      "step": 15360
+    },
+    {
+      "epoch": 46.16,
+      "grad_norm": 14.004356384277344,
+      "learning_rate": 8.461461461461463e-06,
+      "loss": 0.6887,
+      "step": 15370
+    },
+    {
+      "epoch": 46.19,
+      "grad_norm": 14.065787315368652,
+      "learning_rate": 8.460460460460461e-06,
+      "loss": 0.6011,
+      "step": 15380
+    },
+    {
+      "epoch": 46.22,
+      "grad_norm": 16.010286331176758,
+      "learning_rate": 8.45945945945946e-06,
+      "loss": 0.6151,
+      "step": 15390
+    },
+    {
+      "epoch": 46.25,
+      "grad_norm": 20.490942001342773,
+      "learning_rate": 8.45845845845846e-06,
+      "loss": 0.6239,
+      "step": 15400
+    },
+    {
+      "epoch": 46.28,
+      "grad_norm": 17.279037475585938,
+      "learning_rate": 8.457457457457459e-06,
+      "loss": 0.6526,
+      "step": 15410
+    },
+    {
+      "epoch": 46.31,
+      "grad_norm": 16.7100830078125,
+      "learning_rate": 8.456456456456457e-06,
+      "loss": 0.6632,
+      "step": 15420
+    },
+    {
+      "epoch": 46.34,
+      "grad_norm": 13.888665199279785,
+      "learning_rate": 8.455455455455456e-06,
+      "loss": 0.6362,
+      "step": 15430
+    },
+    {
+      "epoch": 46.37,
+      "grad_norm": 13.176531791687012,
+      "learning_rate": 8.454454454454455e-06,
+      "loss": 0.6445,
+      "step": 15440
+    },
+    {
+      "epoch": 46.4,
+      "grad_norm": 25.23253631591797,
+      "learning_rate": 8.453453453453453e-06,
+      "loss": 0.5765,
+      "step": 15450
+    },
+    {
+      "epoch": 46.43,
+      "grad_norm": 14.532776832580566,
+      "learning_rate": 8.452452452452454e-06,
+      "loss": 0.6501,
+      "step": 15460
+    },
+    {
+      "epoch": 46.46,
+      "grad_norm": 15.153827667236328,
+      "learning_rate": 8.451451451451452e-06,
+      "loss": 0.6621,
+      "step": 15470
+    },
+    {
+      "epoch": 46.49,
+      "grad_norm": 14.688456535339355,
+      "learning_rate": 8.45045045045045e-06,
+      "loss": 0.6491,
+      "step": 15480
+    },
+    {
+      "epoch": 46.52,
+      "grad_norm": 18.969226837158203,
+      "learning_rate": 8.449449449449451e-06,
+      "loss": 0.6258,
+      "step": 15490
+    },
+    {
+      "epoch": 46.55,
+      "grad_norm": 16.73055076599121,
+      "learning_rate": 8.44844844844845e-06,
+      "loss": 0.6336,
+      "step": 15500
+    },
+    {
+      "epoch": 46.58,
+      "grad_norm": 12.873644828796387,
+      "learning_rate": 8.447447447447448e-06,
+      "loss": 0.6116,
+      "step": 15510
+    },
+    {
+      "epoch": 46.61,
+      "grad_norm": 15.362987518310547,
+      "learning_rate": 8.446446446446447e-06,
+      "loss": 0.6924,
+      "step": 15520
+    },
+    {
+      "epoch": 46.64,
+      "grad_norm": 15.47192096710205,
+      "learning_rate": 8.445445445445445e-06,
+      "loss": 0.6528,
+      "step": 15530
+    },
+    {
+      "epoch": 46.67,
+      "grad_norm": 20.89593505859375,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 0.6524,
+      "step": 15540
+    },
+    {
+      "epoch": 46.7,
+      "grad_norm": 24.830190658569336,
+      "learning_rate": 8.443443443443444e-06,
+      "loss": 0.6957,
+      "step": 15550
+    },
+    {
+      "epoch": 46.73,
+      "grad_norm": 17.25192642211914,
+      "learning_rate": 8.442442442442443e-06,
+      "loss": 0.6185,
+      "step": 15560
+    },
+    {
+      "epoch": 46.76,
+      "grad_norm": 17.689939498901367,
+      "learning_rate": 8.441441441441443e-06,
+      "loss": 0.6935,
+      "step": 15570
+    },
+    {
+      "epoch": 46.79,
+      "grad_norm": 13.986361503601074,
+      "learning_rate": 8.440440440440442e-06,
+      "loss": 0.6663,
+      "step": 15580
+    },
+    {
+      "epoch": 46.82,
+      "grad_norm": 16.990419387817383,
+      "learning_rate": 8.43943943943944e-06,
+      "loss": 0.6245,
+      "step": 15590
+    },
+    {
+      "epoch": 46.85,
+      "grad_norm": 15.792227745056152,
+      "learning_rate": 8.438438438438439e-06,
+      "loss": 0.7197,
+      "step": 15600
+    },
+    {
+      "epoch": 46.88,
+      "grad_norm": 13.95806884765625,
+      "learning_rate": 8.437437437437438e-06,
+      "loss": 0.6644,
+      "step": 15610
+    },
+    {
+      "epoch": 46.91,
+      "grad_norm": 17.619291305541992,
+      "learning_rate": 8.436436436436438e-06,
+      "loss": 0.6191,
+      "step": 15620
+    },
+    {
+      "epoch": 46.94,
+      "grad_norm": 18.768346786499023,
+      "learning_rate": 8.435435435435436e-06,
+      "loss": 0.6689,
+      "step": 15630
+    },
+    {
+      "epoch": 46.97,
+      "grad_norm": 15.373311042785645,
+      "learning_rate": 8.434434434434435e-06,
+      "loss": 0.6363,
+      "step": 15640
+    },
+    {
+      "epoch": 47.0,
+      "grad_norm": 14.394283294677734,
+      "learning_rate": 8.433433433433435e-06,
+      "loss": 0.6972,
+      "step": 15650
+    },
+    {
+      "epoch": 47.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.44944092631340027,
+      "eval_runtime": 12.6296,
+      "eval_samples_per_second": 791.789,
+      "eval_steps_per_second": 3.167,
+      "step": 15651
+    },
+    {
+      "epoch": 47.03,
+      "grad_norm": 20.598209381103516,
+      "learning_rate": 8.432432432432434e-06,
+      "loss": 0.7864,
+      "step": 15660
+    },
+    {
+      "epoch": 47.06,
+      "grad_norm": 13.101500511169434,
+      "learning_rate": 8.43143143143143e-06,
+      "loss": 0.6511,
+      "step": 15670
+    },
+    {
+      "epoch": 47.09,
+      "grad_norm": 14.161370277404785,
+      "learning_rate": 8.430430430430431e-06,
+      "loss": 0.6281,
+      "step": 15680
+    },
+    {
+      "epoch": 47.12,
+      "grad_norm": 15.656643867492676,
+      "learning_rate": 8.42942942942943e-06,
+      "loss": 0.6166,
+      "step": 15690
+    },
+    {
+      "epoch": 47.15,
+      "grad_norm": 14.171300888061523,
+      "learning_rate": 8.428428428428428e-06,
+      "loss": 0.7045,
+      "step": 15700
+    },
+    {
+      "epoch": 47.18,
+      "grad_norm": 14.199458122253418,
+      "learning_rate": 8.427427427427429e-06,
+      "loss": 0.6232,
+      "step": 15710
+    },
+    {
+      "epoch": 47.21,
+      "grad_norm": 15.500053405761719,
+      "learning_rate": 8.426426426426427e-06,
+      "loss": 0.6978,
+      "step": 15720
+    },
+    {
+      "epoch": 47.24,
+      "grad_norm": 14.61178970336914,
+      "learning_rate": 8.425425425425426e-06,
+      "loss": 0.7303,
+      "step": 15730
+    },
+    {
+      "epoch": 47.27,
+      "grad_norm": 20.73687171936035,
+      "learning_rate": 8.424424424424426e-06,
+      "loss": 0.6623,
+      "step": 15740
+    },
+    {
+      "epoch": 47.3,
+      "grad_norm": 12.876822471618652,
+      "learning_rate": 8.423423423423423e-06,
+      "loss": 0.6474,
+      "step": 15750
+    },
+    {
+      "epoch": 47.33,
+      "grad_norm": 11.815642356872559,
+      "learning_rate": 8.422422422422423e-06,
+      "loss": 0.6512,
+      "step": 15760
+    },
+    {
+      "epoch": 47.36,
+      "grad_norm": 13.063693046569824,
+      "learning_rate": 8.421421421421422e-06,
+      "loss": 0.7295,
+      "step": 15770
+    },
+    {
+      "epoch": 47.39,
+      "grad_norm": 10.594338417053223,
+      "learning_rate": 8.42042042042042e-06,
+      "loss": 0.6597,
+      "step": 15780
+    },
+    {
+      "epoch": 47.42,
+      "grad_norm": 19.250690460205078,
+      "learning_rate": 8.41941941941942e-06,
+      "loss": 0.6416,
+      "step": 15790
+    },
+    {
+      "epoch": 47.45,
+      "grad_norm": 13.238747596740723,
+      "learning_rate": 8.41841841841842e-06,
+      "loss": 0.6216,
+      "step": 15800
+    },
+    {
+      "epoch": 47.48,
+      "grad_norm": 16.948776245117188,
+      "learning_rate": 8.417417417417418e-06,
+      "loss": 0.6794,
+      "step": 15810
+    },
+    {
+      "epoch": 47.51,
+      "grad_norm": 17.4051570892334,
+      "learning_rate": 8.416416416416418e-06,
+      "loss": 0.6369,
+      "step": 15820
+    },
+    {
+      "epoch": 47.54,
+      "grad_norm": 9.287778854370117,
+      "learning_rate": 8.415415415415417e-06,
+      "loss": 0.6962,
+      "step": 15830
+    },
+    {
+      "epoch": 47.57,
+      "grad_norm": 17.206253051757812,
+      "learning_rate": 8.414414414414415e-06,
+      "loss": 0.6219,
+      "step": 15840
+    },
+    {
+      "epoch": 47.6,
+      "grad_norm": 18.415864944458008,
+      "learning_rate": 8.413413413413414e-06,
+      "loss": 0.6386,
+      "step": 15850
+    },
+    {
+      "epoch": 47.63,
+      "grad_norm": 18.562761306762695,
+      "learning_rate": 8.412412412412413e-06,
+      "loss": 0.6297,
+      "step": 15860
+    },
+    {
+      "epoch": 47.66,
+      "grad_norm": 11.125750541687012,
+      "learning_rate": 8.411411411411413e-06,
+      "loss": 0.6703,
+      "step": 15870
+    },
+    {
+      "epoch": 47.69,
+      "grad_norm": 12.335123062133789,
+      "learning_rate": 8.410410410410411e-06,
+      "loss": 0.6603,
+      "step": 15880
+    },
+    {
+      "epoch": 47.72,
+      "grad_norm": 13.002345085144043,
+      "learning_rate": 8.40940940940941e-06,
+      "loss": 0.6128,
+      "step": 15890
+    },
+    {
+      "epoch": 47.75,
+      "grad_norm": 24.356122970581055,
+      "learning_rate": 8.408408408408409e-06,
+      "loss": 0.5832,
+      "step": 15900
+    },
+    {
+      "epoch": 47.78,
+      "grad_norm": 11.936380386352539,
+      "learning_rate": 8.407407407407409e-06,
+      "loss": 0.6503,
+      "step": 15910
+    },
+    {
+      "epoch": 47.81,
+      "grad_norm": 16.61817741394043,
+      "learning_rate": 8.406406406406406e-06,
+      "loss": 0.6221,
+      "step": 15920
+    },
+    {
+      "epoch": 47.84,
+      "grad_norm": 17.528955459594727,
+      "learning_rate": 8.405405405405406e-06,
+      "loss": 0.6718,
+      "step": 15930
+    },
+    {
+      "epoch": 47.87,
+      "grad_norm": 14.016353607177734,
+      "learning_rate": 8.404404404404405e-06,
+      "loss": 0.627,
+      "step": 15940
+    },
+    {
+      "epoch": 47.9,
+      "grad_norm": 9.69521427154541,
+      "learning_rate": 8.403403403403403e-06,
+      "loss": 0.6385,
+      "step": 15950
+    },
+    {
+      "epoch": 47.93,
+      "grad_norm": 17.623855590820312,
+      "learning_rate": 8.402402402402404e-06,
+      "loss": 0.6143,
+      "step": 15960
+    },
+    {
+      "epoch": 47.96,
+      "grad_norm": 12.067631721496582,
+      "learning_rate": 8.401401401401402e-06,
+      "loss": 0.6262,
+      "step": 15970
+    },
+    {
+      "epoch": 47.99,
+      "grad_norm": 13.75415325164795,
+      "learning_rate": 8.4004004004004e-06,
+      "loss": 0.616,
+      "step": 15980
+    },
+    {
+      "epoch": 48.0,
+      "eval_accuracy": 0.8674,
+      "eval_loss": 0.45251065492630005,
+      "eval_runtime": 12.7377,
+      "eval_samples_per_second": 785.071,
+      "eval_steps_per_second": 3.14,
+      "step": 15984
+    },
+    {
+      "epoch": 48.02,
+      "grad_norm": 14.465485572814941,
+      "learning_rate": 8.399399399399401e-06,
+      "loss": 0.5542,
+      "step": 15990
+    },
+    {
+      "epoch": 48.05,
+      "grad_norm": 14.680464744567871,
+      "learning_rate": 8.398398398398398e-06,
+      "loss": 0.7044,
+      "step": 16000
+    },
+    {
+      "epoch": 48.08,
+      "grad_norm": 11.977607727050781,
+      "learning_rate": 8.397397397397398e-06,
+      "loss": 0.6151,
+      "step": 16010
+    },
+    {
+      "epoch": 48.11,
+      "grad_norm": 17.933095932006836,
+      "learning_rate": 8.396396396396397e-06,
+      "loss": 0.6358,
+      "step": 16020
+    },
+    {
+      "epoch": 48.14,
+      "grad_norm": 15.633028984069824,
+      "learning_rate": 8.395395395395395e-06,
+      "loss": 0.657,
+      "step": 16030
+    },
+    {
+      "epoch": 48.17,
+      "grad_norm": 14.462199211120605,
+      "learning_rate": 8.394394394394396e-06,
+      "loss": 0.6752,
+      "step": 16040
+    },
+    {
+      "epoch": 48.2,
+      "grad_norm": 12.743391990661621,
+      "learning_rate": 8.393393393393394e-06,
+      "loss": 0.6515,
+      "step": 16050
+    },
+    {
+      "epoch": 48.23,
+      "grad_norm": 12.953666687011719,
+      "learning_rate": 8.392392392392393e-06,
+      "loss": 0.6412,
+      "step": 16060
+    },
+    {
+      "epoch": 48.26,
+      "grad_norm": 14.461003303527832,
+      "learning_rate": 8.391391391391393e-06,
+      "loss": 0.6275,
+      "step": 16070
+    },
+    {
+      "epoch": 48.29,
+      "grad_norm": 18.0173397064209,
+      "learning_rate": 8.390390390390392e-06,
+      "loss": 0.7052,
+      "step": 16080
+    },
+    {
+      "epoch": 48.32,
+      "grad_norm": 13.393255233764648,
+      "learning_rate": 8.38938938938939e-06,
+      "loss": 0.6697,
+      "step": 16090
+    },
+    {
+      "epoch": 48.35,
+      "grad_norm": 14.54150676727295,
+      "learning_rate": 8.388388388388389e-06,
+      "loss": 0.6551,
+      "step": 16100
+    },
+    {
+      "epoch": 48.38,
+      "grad_norm": 16.1068115234375,
+      "learning_rate": 8.387387387387388e-06,
+      "loss": 0.6035,
+      "step": 16110
+    },
+    {
+      "epoch": 48.41,
+      "grad_norm": 14.576680183410645,
+      "learning_rate": 8.386386386386386e-06,
+      "loss": 0.6353,
+      "step": 16120
+    },
+    {
+      "epoch": 48.44,
+      "grad_norm": 20.142234802246094,
+      "learning_rate": 8.385385385385386e-06,
+      "loss": 0.6536,
+      "step": 16130
+    },
+    {
+      "epoch": 48.47,
+      "grad_norm": 14.841194152832031,
+      "learning_rate": 8.384384384384385e-06,
+      "loss": 0.6272,
+      "step": 16140
+    },
+    {
+      "epoch": 48.5,
+      "grad_norm": 15.862102508544922,
+      "learning_rate": 8.383383383383384e-06,
+      "loss": 0.5461,
+      "step": 16150
+    },
+    {
+      "epoch": 48.53,
+      "grad_norm": 22.324954986572266,
+      "learning_rate": 8.382382382382384e-06,
+      "loss": 0.6802,
+      "step": 16160
+    },
+    {
+      "epoch": 48.56,
+      "grad_norm": 29.10227394104004,
+      "learning_rate": 8.381381381381381e-06,
+      "loss": 0.6632,
+      "step": 16170
+    },
+    {
+      "epoch": 48.59,
+      "grad_norm": 15.7824125289917,
+      "learning_rate": 8.380380380380381e-06,
+      "loss": 0.6967,
+      "step": 16180
+    },
+    {
+      "epoch": 48.62,
+      "grad_norm": 17.80396842956543,
+      "learning_rate": 8.37937937937938e-06,
+      "loss": 0.6012,
+      "step": 16190
+    },
+    {
+      "epoch": 48.65,
+      "grad_norm": 17.60359764099121,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.6098,
+      "step": 16200
+    },
+    {
+      "epoch": 48.68,
+      "grad_norm": 13.17934799194336,
+      "learning_rate": 8.377377377377379e-06,
+      "loss": 0.6558,
+      "step": 16210
+    },
+    {
+      "epoch": 48.71,
+      "grad_norm": 18.207441329956055,
+      "learning_rate": 8.376376376376377e-06,
+      "loss": 0.6079,
+      "step": 16220
+    },
+    {
+      "epoch": 48.74,
+      "grad_norm": 13.500203132629395,
+      "learning_rate": 8.375375375375376e-06,
+      "loss": 0.6025,
+      "step": 16230
+    },
+    {
+      "epoch": 48.77,
+      "grad_norm": 13.301795959472656,
+      "learning_rate": 8.374374374374376e-06,
+      "loss": 0.6264,
+      "step": 16240
+    },
+    {
+      "epoch": 48.8,
+      "grad_norm": 14.00633430480957,
+      "learning_rate": 8.373373373373373e-06,
+      "loss": 0.7114,
+      "step": 16250
+    },
+    {
+      "epoch": 48.83,
+      "grad_norm": 15.595272064208984,
+      "learning_rate": 8.372372372372373e-06,
+      "loss": 0.5517,
+      "step": 16260
+    },
+    {
+      "epoch": 48.86,
+      "grad_norm": 16.70868492126465,
+      "learning_rate": 8.371371371371372e-06,
+      "loss": 0.6613,
+      "step": 16270
+    },
+    {
+      "epoch": 48.89,
+      "grad_norm": 14.774352073669434,
+      "learning_rate": 8.37037037037037e-06,
+      "loss": 0.5916,
+      "step": 16280
+    },
+    {
+      "epoch": 48.92,
+      "grad_norm": 13.337324142456055,
+      "learning_rate": 8.36936936936937e-06,
+      "loss": 0.6354,
+      "step": 16290
+    },
+    {
+      "epoch": 48.95,
+      "grad_norm": 16.758169174194336,
+      "learning_rate": 8.36836836836837e-06,
+      "loss": 0.6615,
+      "step": 16300
+    },
+    {
+      "epoch": 48.98,
+      "grad_norm": 17.271648406982422,
+      "learning_rate": 8.367367367367368e-06,
+      "loss": 0.6911,
+      "step": 16310
+    },
+    {
+      "epoch": 49.0,
+      "eval_accuracy": 0.8637,
+      "eval_loss": 0.4506002366542816,
+      "eval_runtime": 12.6548,
+      "eval_samples_per_second": 790.216,
+      "eval_steps_per_second": 3.161,
+      "step": 16317
+    },
+    {
+      "epoch": 49.01,
+      "grad_norm": 16.090843200683594,
+      "learning_rate": 8.366366366366368e-06,
+      "loss": 0.7275,
+      "step": 16320
+    },
+    {
+      "epoch": 49.04,
+      "grad_norm": 15.7288179397583,
+      "learning_rate": 8.365365365365367e-06,
+      "loss": 0.6623,
+      "step": 16330
+    },
+    {
+      "epoch": 49.07,
+      "grad_norm": 19.782133102416992,
+      "learning_rate": 8.364364364364365e-06,
+      "loss": 0.5901,
+      "step": 16340
+    },
+    {
+      "epoch": 49.1,
+      "grad_norm": 19.402259826660156,
+      "learning_rate": 8.363363363363364e-06,
+      "loss": 0.5984,
+      "step": 16350
+    },
+    {
+      "epoch": 49.13,
+      "grad_norm": 12.4588041305542,
+      "learning_rate": 8.362362362362363e-06,
+      "loss": 0.6633,
+      "step": 16360
+    },
+    {
+      "epoch": 49.16,
+      "grad_norm": 13.56613826751709,
+      "learning_rate": 8.361361361361361e-06,
+      "loss": 0.6799,
+      "step": 16370
+    },
+    {
+      "epoch": 49.19,
+      "grad_norm": 17.75737762451172,
+      "learning_rate": 8.360360360360362e-06,
+      "loss": 0.5668,
+      "step": 16380
+    },
+    {
+      "epoch": 49.22,
+      "grad_norm": 20.319490432739258,
+      "learning_rate": 8.35935935935936e-06,
+      "loss": 0.6673,
+      "step": 16390
+    },
+    {
+      "epoch": 49.25,
+      "grad_norm": 28.126279830932617,
+      "learning_rate": 8.358358358358359e-06,
+      "loss": 0.6171,
+      "step": 16400
+    },
+    {
+      "epoch": 49.28,
+      "grad_norm": 17.8362979888916,
+      "learning_rate": 8.357357357357359e-06,
+      "loss": 0.619,
+      "step": 16410
+    },
+    {
+      "epoch": 49.31,
+      "grad_norm": 15.377138137817383,
+      "learning_rate": 8.356356356356356e-06,
+      "loss": 0.6238,
+      "step": 16420
+    },
+    {
+      "epoch": 49.34,
+      "grad_norm": 13.098530769348145,
+      "learning_rate": 8.355355355355356e-06,
+      "loss": 0.6247,
+      "step": 16430
+    },
+    {
+      "epoch": 49.37,
+      "grad_norm": 25.9362850189209,
+      "learning_rate": 8.354354354354355e-06,
+      "loss": 0.6544,
+      "step": 16440
+    },
+    {
+      "epoch": 49.4,
+      "grad_norm": 14.757615089416504,
+      "learning_rate": 8.353353353353353e-06,
+      "loss": 0.6629,
+      "step": 16450
+    },
+    {
+      "epoch": 49.43,
+      "grad_norm": 16.951858520507812,
+      "learning_rate": 8.352352352352354e-06,
+      "loss": 0.6348,
+      "step": 16460
+    },
+    {
+      "epoch": 49.46,
+      "grad_norm": 20.39686393737793,
+      "learning_rate": 8.351351351351352e-06,
+      "loss": 0.7222,
+      "step": 16470
+    },
+    {
+      "epoch": 49.49,
+      "grad_norm": 11.490650177001953,
+      "learning_rate": 8.35035035035035e-06,
+      "loss": 0.5808,
+      "step": 16480
+    },
+    {
+      "epoch": 49.52,
+      "grad_norm": 16.503400802612305,
+      "learning_rate": 8.349349349349351e-06,
+      "loss": 0.6684,
+      "step": 16490
+    },
+    {
+      "epoch": 49.55,
+      "grad_norm": 10.76430606842041,
+      "learning_rate": 8.348348348348348e-06,
+      "loss": 0.6216,
+      "step": 16500
+    },
+    {
+      "epoch": 49.58,
+      "grad_norm": 18.0540771484375,
+      "learning_rate": 8.347347347347348e-06,
+      "loss": 0.6222,
+      "step": 16510
+    },
+    {
+      "epoch": 49.61,
+      "grad_norm": 26.5042781829834,
+      "learning_rate": 8.346346346346347e-06,
+      "loss": 0.6964,
+      "step": 16520
+    },
+    {
+      "epoch": 49.64,
+      "grad_norm": 21.34775161743164,
+      "learning_rate": 8.345345345345346e-06,
+      "loss": 0.6164,
+      "step": 16530
+    },
+    {
+      "epoch": 49.67,
+      "grad_norm": 14.51412582397461,
+      "learning_rate": 8.344344344344346e-06,
+      "loss": 0.6464,
+      "step": 16540
+    },
+    {
+      "epoch": 49.7,
+      "grad_norm": 14.07699966430664,
+      "learning_rate": 8.343343343343344e-06,
+      "loss": 0.667,
+      "step": 16550
+    },
+    {
+      "epoch": 49.73,
+      "grad_norm": 13.938491821289062,
+      "learning_rate": 8.342342342342343e-06,
+      "loss": 0.6875,
+      "step": 16560
+    },
+    {
+      "epoch": 49.76,
+      "grad_norm": 12.678507804870605,
+      "learning_rate": 8.341341341341343e-06,
+      "loss": 0.5823,
+      "step": 16570
+    },
+    {
+      "epoch": 49.79,
+      "grad_norm": 13.337613105773926,
+      "learning_rate": 8.340340340340342e-06,
+      "loss": 0.6212,
+      "step": 16580
+    },
+    {
+      "epoch": 49.82,
+      "grad_norm": 17.60381317138672,
+      "learning_rate": 8.339339339339339e-06,
+      "loss": 0.6683,
+      "step": 16590
+    },
+    {
+      "epoch": 49.85,
+      "grad_norm": 15.319842338562012,
+      "learning_rate": 8.338338338338339e-06,
+      "loss": 0.6332,
+      "step": 16600
+    },
+    {
+      "epoch": 49.88,
+      "grad_norm": 21.950565338134766,
+      "learning_rate": 8.337337337337338e-06,
+      "loss": 0.5829,
+      "step": 16610
+    },
+    {
+      "epoch": 49.91,
+      "grad_norm": 16.845312118530273,
+      "learning_rate": 8.336336336336336e-06,
+      "loss": 0.6298,
+      "step": 16620
+    },
+    {
+      "epoch": 49.94,
+      "grad_norm": 16.28782081604004,
+      "learning_rate": 8.335335335335337e-06,
+      "loss": 0.6576,
+      "step": 16630
+    },
+    {
+      "epoch": 49.97,
+      "grad_norm": 13.272744178771973,
+      "learning_rate": 8.334334334334335e-06,
+      "loss": 0.583,
+      "step": 16640
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 80.26773834228516,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.6737,
+      "step": 16650
+    },
+    {
+      "epoch": 50.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.4503735899925232,
+      "eval_runtime": 12.6697,
+      "eval_samples_per_second": 789.285,
+      "eval_steps_per_second": 3.157,
+      "step": 16650
+    },
+    {
+      "epoch": 50.03,
+      "grad_norm": 16.50661849975586,
+      "learning_rate": 8.332332332332334e-06,
+      "loss": 0.6137,
+      "step": 16660
+    },
+    {
+      "epoch": 50.06,
+      "grad_norm": 15.900172233581543,
+      "learning_rate": 8.331331331331331e-06,
+      "loss": 0.6835,
+      "step": 16670
+    },
+    {
+      "epoch": 50.09,
+      "grad_norm": 18.15906524658203,
+      "learning_rate": 8.330330330330331e-06,
+      "loss": 0.6425,
+      "step": 16680
+    },
+    {
+      "epoch": 50.12,
+      "grad_norm": 11.704545021057129,
+      "learning_rate": 8.32932932932933e-06,
+      "loss": 0.5861,
+      "step": 16690
+    },
+    {
+      "epoch": 50.15,
+      "grad_norm": 11.482316970825195,
+      "learning_rate": 8.328328328328328e-06,
+      "loss": 0.6083,
+      "step": 16700
+    },
+    {
+      "epoch": 50.18,
+      "grad_norm": 21.32619857788086,
+      "learning_rate": 8.327327327327329e-06,
+      "loss": 0.6729,
+      "step": 16710
+    },
+    {
+      "epoch": 50.21,
+      "grad_norm": 15.120251655578613,
+      "learning_rate": 8.326326326326327e-06,
+      "loss": 0.6147,
+      "step": 16720
+    },
+    {
+      "epoch": 50.24,
+      "grad_norm": 12.25693130493164,
+      "learning_rate": 8.325325325325326e-06,
+      "loss": 0.6118,
+      "step": 16730
+    },
+    {
+      "epoch": 50.27,
+      "grad_norm": 14.365169525146484,
+      "learning_rate": 8.324324324324326e-06,
+      "loss": 0.575,
+      "step": 16740
+    },
+    {
+      "epoch": 50.3,
+      "grad_norm": 16.102943420410156,
+      "learning_rate": 8.323323323323323e-06,
+      "loss": 0.5915,
+      "step": 16750
+    },
+    {
+      "epoch": 50.33,
+      "grad_norm": 16.202205657958984,
+      "learning_rate": 8.322322322322323e-06,
+      "loss": 0.6382,
+      "step": 16760
+    },
+    {
+      "epoch": 50.36,
+      "grad_norm": 15.920069694519043,
+      "learning_rate": 8.321321321321322e-06,
+      "loss": 0.6673,
+      "step": 16770
+    },
+    {
+      "epoch": 50.39,
+      "grad_norm": 14.586955070495605,
+      "learning_rate": 8.32032032032032e-06,
+      "loss": 0.631,
+      "step": 16780
+    },
+    {
+      "epoch": 50.42,
+      "grad_norm": 25.971364974975586,
+      "learning_rate": 8.31931931931932e-06,
+      "loss": 0.5962,
+      "step": 16790
+    },
+    {
+      "epoch": 50.45,
+      "grad_norm": 21.708187103271484,
+      "learning_rate": 8.31831831831832e-06,
+      "loss": 0.5806,
+      "step": 16800
+    },
+    {
+      "epoch": 50.48,
+      "grad_norm": 17.089094161987305,
+      "learning_rate": 8.317317317317318e-06,
+      "loss": 0.6518,
+      "step": 16810
+    },
+    {
+      "epoch": 50.51,
+      "grad_norm": 12.237706184387207,
+      "learning_rate": 8.316316316316317e-06,
+      "loss": 0.6206,
+      "step": 16820
+    },
+    {
+      "epoch": 50.54,
+      "grad_norm": 12.970582008361816,
+      "learning_rate": 8.315315315315317e-06,
+      "loss": 0.5751,
+      "step": 16830
+    },
+    {
+      "epoch": 50.57,
+      "grad_norm": 16.444976806640625,
+      "learning_rate": 8.314314314314314e-06,
+      "loss": 0.5381,
+      "step": 16840
+    },
+    {
+      "epoch": 50.6,
+      "grad_norm": 17.726957321166992,
+      "learning_rate": 8.313313313313314e-06,
+      "loss": 0.6321,
+      "step": 16850
+    },
+    {
+      "epoch": 50.63,
+      "grad_norm": 14.764182090759277,
+      "learning_rate": 8.312312312312313e-06,
+      "loss": 0.6045,
+      "step": 16860
+    },
+    {
+      "epoch": 50.66,
+      "grad_norm": 11.397544860839844,
+      "learning_rate": 8.311311311311311e-06,
+      "loss": 0.6195,
+      "step": 16870
+    },
+    {
+      "epoch": 50.69,
+      "grad_norm": 18.760087966918945,
+      "learning_rate": 8.310310310310312e-06,
+      "loss": 0.6775,
+      "step": 16880
+    },
+    {
+      "epoch": 50.72,
+      "grad_norm": 16.329402923583984,
+      "learning_rate": 8.30930930930931e-06,
+      "loss": 0.5952,
+      "step": 16890
+    },
+    {
+      "epoch": 50.75,
+      "grad_norm": 17.519657135009766,
+      "learning_rate": 8.308308308308309e-06,
+      "loss": 0.6568,
+      "step": 16900
+    },
+    {
+      "epoch": 50.78,
+      "grad_norm": 13.098520278930664,
+      "learning_rate": 8.307307307307309e-06,
+      "loss": 0.6262,
+      "step": 16910
+    },
+    {
+      "epoch": 50.81,
+      "grad_norm": 18.580644607543945,
+      "learning_rate": 8.306306306306306e-06,
+      "loss": 0.6059,
+      "step": 16920
+    },
+    {
+      "epoch": 50.84,
+      "grad_norm": 13.884407043457031,
+      "learning_rate": 8.305305305305306e-06,
+      "loss": 0.646,
+      "step": 16930
+    },
+    {
+      "epoch": 50.87,
+      "grad_norm": 16.32083511352539,
+      "learning_rate": 8.304304304304305e-06,
+      "loss": 0.6617,
+      "step": 16940
+    },
+    {
+      "epoch": 50.9,
+      "grad_norm": 17.042152404785156,
+      "learning_rate": 8.303303303303303e-06,
+      "loss": 0.6276,
+      "step": 16950
+    },
+    {
+      "epoch": 50.93,
+      "grad_norm": 19.203786849975586,
+      "learning_rate": 8.302302302302304e-06,
+      "loss": 0.6461,
+      "step": 16960
+    },
+    {
+      "epoch": 50.96,
+      "grad_norm": 17.717748641967773,
+      "learning_rate": 8.301301301301302e-06,
+      "loss": 0.6799,
+      "step": 16970
+    },
+    {
+      "epoch": 50.99,
+      "grad_norm": 15.912696838378906,
+      "learning_rate": 8.300300300300301e-06,
+      "loss": 0.5573,
+      "step": 16980
+    },
+    {
+      "epoch": 51.0,
+      "eval_accuracy": 0.8641,
+      "eval_loss": 0.4541584551334381,
+      "eval_runtime": 12.9427,
+      "eval_samples_per_second": 772.638,
+      "eval_steps_per_second": 3.091,
+      "step": 16983
+    },
+    {
+      "epoch": 51.02,
+      "grad_norm": 16.83027458190918,
+      "learning_rate": 8.299299299299301e-06,
+      "loss": 0.5235,
+      "step": 16990
+    },
+    {
+      "epoch": 51.05,
+      "grad_norm": 24.17321014404297,
+      "learning_rate": 8.298298298298298e-06,
+      "loss": 0.6485,
+      "step": 17000
+    },
+    {
+      "epoch": 51.08,
+      "grad_norm": 15.718440055847168,
+      "learning_rate": 8.297297297297298e-06,
+      "loss": 0.5895,
+      "step": 17010
+    },
+    {
+      "epoch": 51.11,
+      "grad_norm": 26.84842300415039,
+      "learning_rate": 8.296296296296297e-06,
+      "loss": 0.6912,
+      "step": 17020
+    },
+    {
+      "epoch": 51.14,
+      "grad_norm": 12.315144538879395,
+      "learning_rate": 8.295295295295296e-06,
+      "loss": 0.6454,
+      "step": 17030
+    },
+    {
+      "epoch": 51.17,
+      "grad_norm": 15.44035816192627,
+      "learning_rate": 8.294294294294294e-06,
+      "loss": 0.6382,
+      "step": 17040
+    },
+    {
+      "epoch": 51.2,
+      "grad_norm": 19.862279891967773,
+      "learning_rate": 8.293293293293294e-06,
+      "loss": 0.6557,
+      "step": 17050
+    },
+    {
+      "epoch": 51.23,
+      "grad_norm": 17.58277130126953,
+      "learning_rate": 8.292292292292293e-06,
+      "loss": 0.6643,
+      "step": 17060
+    },
+    {
+      "epoch": 51.26,
+      "grad_norm": 15.316263198852539,
+      "learning_rate": 8.291291291291292e-06,
+      "loss": 0.6296,
+      "step": 17070
+    },
+    {
+      "epoch": 51.29,
+      "grad_norm": 13.260090827941895,
+      "learning_rate": 8.29029029029029e-06,
+      "loss": 0.6293,
+      "step": 17080
+    },
+    {
+      "epoch": 51.32,
+      "grad_norm": 18.472536087036133,
+      "learning_rate": 8.289289289289289e-06,
+      "loss": 0.6279,
+      "step": 17090
+    },
+    {
+      "epoch": 51.35,
+      "grad_norm": 18.13410758972168,
+      "learning_rate": 8.288288288288289e-06,
+      "loss": 0.5917,
+      "step": 17100
+    },
+    {
+      "epoch": 51.38,
+      "grad_norm": 12.870095252990723,
+      "learning_rate": 8.287287287287288e-06,
+      "loss": 0.5901,
+      "step": 17110
+    },
+    {
+      "epoch": 51.41,
+      "grad_norm": 13.40429973602295,
+      "learning_rate": 8.286286286286286e-06,
+      "loss": 0.5984,
+      "step": 17120
+    },
+    {
+      "epoch": 51.44,
+      "grad_norm": 19.12213134765625,
+      "learning_rate": 8.285285285285287e-06,
+      "loss": 0.6627,
+      "step": 17130
+    },
+    {
+      "epoch": 51.47,
+      "grad_norm": 12.901103973388672,
+      "learning_rate": 8.284284284284285e-06,
+      "loss": 0.6143,
+      "step": 17140
+    },
+    {
+      "epoch": 51.5,
+      "grad_norm": 18.621458053588867,
+      "learning_rate": 8.283283283283284e-06,
+      "loss": 0.5618,
+      "step": 17150
+    },
+    {
+      "epoch": 51.53,
+      "grad_norm": 17.519861221313477,
+      "learning_rate": 8.282282282282284e-06,
+      "loss": 0.6703,
+      "step": 17160
+    },
+    {
+      "epoch": 51.56,
+      "grad_norm": 23.81211280822754,
+      "learning_rate": 8.281281281281281e-06,
+      "loss": 0.6219,
+      "step": 17170
+    },
+    {
+      "epoch": 51.59,
+      "grad_norm": 16.727157592773438,
+      "learning_rate": 8.280280280280281e-06,
+      "loss": 0.5719,
+      "step": 17180
+    },
+    {
+      "epoch": 51.62,
+      "grad_norm": 15.704798698425293,
+      "learning_rate": 8.27927927927928e-06,
+      "loss": 0.6613,
+      "step": 17190
+    },
+    {
+      "epoch": 51.65,
+      "grad_norm": 20.87688636779785,
+      "learning_rate": 8.278278278278278e-06,
+      "loss": 0.5979,
+      "step": 17200
+    },
+    {
+      "epoch": 51.68,
+      "grad_norm": 14.883535385131836,
+      "learning_rate": 8.277277277277279e-06,
+      "loss": 0.6128,
+      "step": 17210
+    },
+    {
+      "epoch": 51.71,
+      "grad_norm": 18.046480178833008,
+      "learning_rate": 8.276276276276277e-06,
+      "loss": 0.6295,
+      "step": 17220
+    },
+    {
+      "epoch": 51.74,
+      "grad_norm": 14.467040061950684,
+      "learning_rate": 8.275275275275276e-06,
+      "loss": 0.6666,
+      "step": 17230
+    },
+    {
+      "epoch": 51.77,
+      "grad_norm": 13.173455238342285,
+      "learning_rate": 8.274274274274276e-06,
+      "loss": 0.5738,
+      "step": 17240
+    },
+    {
+      "epoch": 51.8,
+      "grad_norm": 15.290773391723633,
+      "learning_rate": 8.273273273273273e-06,
+      "loss": 0.6693,
+      "step": 17250
+    },
+    {
+      "epoch": 51.83,
+      "grad_norm": 15.725456237792969,
+      "learning_rate": 8.272272272272273e-06,
+      "loss": 0.6455,
+      "step": 17260
+    },
+    {
+      "epoch": 51.86,
+      "grad_norm": 19.449800491333008,
+      "learning_rate": 8.271271271271272e-06,
+      "loss": 0.6245,
+      "step": 17270
+    },
+    {
+      "epoch": 51.89,
+      "grad_norm": 15.260242462158203,
+      "learning_rate": 8.27027027027027e-06,
+      "loss": 0.5877,
+      "step": 17280
+    },
+    {
+      "epoch": 51.92,
+      "grad_norm": 16.413896560668945,
+      "learning_rate": 8.26926926926927e-06,
+      "loss": 0.6061,
+      "step": 17290
+    },
+    {
+      "epoch": 51.95,
+      "grad_norm": 12.333784103393555,
+      "learning_rate": 8.26826826826827e-06,
+      "loss": 0.6318,
+      "step": 17300
+    },
+    {
+      "epoch": 51.98,
+      "grad_norm": 17.20624351501465,
+      "learning_rate": 8.267267267267268e-06,
+      "loss": 0.6296,
+      "step": 17310
+    },
+    {
+      "epoch": 52.0,
+      "eval_accuracy": 0.8626,
+      "eval_loss": 0.45732319355010986,
+      "eval_runtime": 13.213,
+      "eval_samples_per_second": 756.831,
+      "eval_steps_per_second": 3.027,
+      "step": 17316
+    },
+    {
+      "epoch": 52.01,
+      "grad_norm": 15.85933780670166,
+      "learning_rate": 8.266266266266267e-06,
+      "loss": 0.6516,
+      "step": 17320
+    },
+    {
+      "epoch": 52.04,
+      "grad_norm": 21.219789505004883,
+      "learning_rate": 8.265265265265265e-06,
+      "loss": 0.6241,
+      "step": 17330
+    },
+    {
+      "epoch": 52.07,
+      "grad_norm": 16.367534637451172,
+      "learning_rate": 8.264264264264264e-06,
+      "loss": 0.6824,
+      "step": 17340
+    },
+    {
+      "epoch": 52.1,
+      "grad_norm": 18.358396530151367,
+      "learning_rate": 8.263263263263264e-06,
+      "loss": 0.6016,
+      "step": 17350
+    },
+    {
+      "epoch": 52.13,
+      "grad_norm": 13.449258804321289,
+      "learning_rate": 8.262262262262263e-06,
+      "loss": 0.6292,
+      "step": 17360
+    },
+    {
+      "epoch": 52.16,
+      "grad_norm": 19.812196731567383,
+      "learning_rate": 8.261261261261261e-06,
+      "loss": 0.6062,
+      "step": 17370
+    },
+    {
+      "epoch": 52.19,
+      "grad_norm": 16.552570343017578,
+      "learning_rate": 8.260260260260262e-06,
+      "loss": 0.6181,
+      "step": 17380
+    },
+    {
+      "epoch": 52.22,
+      "grad_norm": 13.700467109680176,
+      "learning_rate": 8.25925925925926e-06,
+      "loss": 0.6256,
+      "step": 17390
+    },
+    {
+      "epoch": 52.25,
+      "grad_norm": 16.087739944458008,
+      "learning_rate": 8.258258258258259e-06,
+      "loss": 0.624,
+      "step": 17400
+    },
+    {
+      "epoch": 52.28,
+      "grad_norm": 13.194904327392578,
+      "learning_rate": 8.257257257257259e-06,
+      "loss": 0.6256,
+      "step": 17410
+    },
+    {
+      "epoch": 52.31,
+      "grad_norm": 20.677940368652344,
+      "learning_rate": 8.256256256256256e-06,
+      "loss": 0.6342,
+      "step": 17420
+    },
+    {
+      "epoch": 52.34,
+      "grad_norm": 17.496768951416016,
+      "learning_rate": 8.255255255255256e-06,
+      "loss": 0.6425,
+      "step": 17430
+    },
+    {
+      "epoch": 52.37,
+      "grad_norm": 16.56483268737793,
+      "learning_rate": 8.254254254254255e-06,
+      "loss": 0.5831,
+      "step": 17440
+    },
+    {
+      "epoch": 52.4,
+      "grad_norm": 20.45583152770996,
+      "learning_rate": 8.253253253253254e-06,
+      "loss": 0.6419,
+      "step": 17450
+    },
+    {
+      "epoch": 52.43,
+      "grad_norm": 15.296815872192383,
+      "learning_rate": 8.252252252252254e-06,
+      "loss": 0.5942,
+      "step": 17460
+    },
+    {
+      "epoch": 52.46,
+      "grad_norm": 13.59698486328125,
+      "learning_rate": 8.251251251251252e-06,
+      "loss": 0.5907,
+      "step": 17470
+    },
+    {
+      "epoch": 52.49,
+      "grad_norm": 15.531094551086426,
+      "learning_rate": 8.250250250250251e-06,
+      "loss": 0.6174,
+      "step": 17480
+    },
+    {
+      "epoch": 52.52,
+      "grad_norm": 11.718438148498535,
+      "learning_rate": 8.24924924924925e-06,
+      "loss": 0.6663,
+      "step": 17490
+    },
+    {
+      "epoch": 52.55,
+      "grad_norm": 13.98644733428955,
+      "learning_rate": 8.248248248248248e-06,
+      "loss": 0.598,
+      "step": 17500
+    },
+    {
+      "epoch": 52.58,
+      "grad_norm": 12.666622161865234,
+      "learning_rate": 8.247247247247247e-06,
+      "loss": 0.6163,
+      "step": 17510
+    },
+    {
+      "epoch": 52.61,
+      "grad_norm": 18.385515213012695,
+      "learning_rate": 8.246246246246247e-06,
+      "loss": 0.6184,
+      "step": 17520
+    },
+    {
+      "epoch": 52.64,
+      "grad_norm": 21.152790069580078,
+      "learning_rate": 8.245245245245246e-06,
+      "loss": 0.6664,
+      "step": 17530
+    },
+    {
+      "epoch": 52.67,
+      "grad_norm": 14.228870391845703,
+      "learning_rate": 8.244244244244244e-06,
+      "loss": 0.6337,
+      "step": 17540
+    },
+    {
+      "epoch": 52.7,
+      "grad_norm": 19.32399559020996,
+      "learning_rate": 8.243243243243245e-06,
+      "loss": 0.6542,
+      "step": 17550
+    },
+    {
+      "epoch": 52.73,
+      "grad_norm": 14.791872024536133,
+      "learning_rate": 8.242242242242243e-06,
+      "loss": 0.6347,
+      "step": 17560
+    },
+    {
+      "epoch": 52.76,
+      "grad_norm": 12.860326766967773,
+      "learning_rate": 8.241241241241242e-06,
+      "loss": 0.5534,
+      "step": 17570
+    },
+    {
+      "epoch": 52.79,
+      "grad_norm": 13.235801696777344,
+      "learning_rate": 8.24024024024024e-06,
+      "loss": 0.6217,
+      "step": 17580
+    },
+    {
+      "epoch": 52.82,
+      "grad_norm": 16.83515739440918,
+      "learning_rate": 8.239239239239239e-06,
+      "loss": 0.6346,
+      "step": 17590
+    },
+    {
+      "epoch": 52.85,
+      "grad_norm": 13.064797401428223,
+      "learning_rate": 8.23823823823824e-06,
+      "loss": 0.6224,
+      "step": 17600
+    },
+    {
+      "epoch": 52.88,
+      "grad_norm": 17.000850677490234,
+      "learning_rate": 8.237237237237238e-06,
+      "loss": 0.6116,
+      "step": 17610
+    },
+    {
+      "epoch": 52.91,
+      "grad_norm": 13.369550704956055,
+      "learning_rate": 8.236236236236236e-06,
+      "loss": 0.6734,
+      "step": 17620
+    },
+    {
+      "epoch": 52.94,
+      "grad_norm": 13.209344863891602,
+      "learning_rate": 8.235235235235237e-06,
+      "loss": 0.6274,
+      "step": 17630
+    },
+    {
+      "epoch": 52.97,
+      "grad_norm": 17.668176651000977,
+      "learning_rate": 8.234234234234235e-06,
+      "loss": 0.6245,
+      "step": 17640
+    },
+    {
+      "epoch": 53.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.45495444536209106,
+      "eval_runtime": 13.0186,
+      "eval_samples_per_second": 768.132,
+      "eval_steps_per_second": 3.073,
+      "step": 17649
+    },
+    {
+      "epoch": 53.0,
+      "grad_norm": 12.670238494873047,
+      "learning_rate": 8.233233233233234e-06,
+      "loss": 0.6691,
+      "step": 17650
+    },
+    {
+      "epoch": 53.03,
+      "grad_norm": 17.616172790527344,
+      "learning_rate": 8.232232232232234e-06,
+      "loss": 0.6252,
+      "step": 17660
+    },
+    {
+      "epoch": 53.06,
+      "grad_norm": 14.60269546508789,
+      "learning_rate": 8.231231231231231e-06,
+      "loss": 0.6346,
+      "step": 17670
+    },
+    {
+      "epoch": 53.09,
+      "grad_norm": 15.650839805603027,
+      "learning_rate": 8.230230230230231e-06,
+      "loss": 0.6162,
+      "step": 17680
+    },
+    {
+      "epoch": 53.12,
+      "grad_norm": 15.538573265075684,
+      "learning_rate": 8.22922922922923e-06,
+      "loss": 0.5013,
+      "step": 17690
+    },
+    {
+      "epoch": 53.15,
+      "grad_norm": 13.48410701751709,
+      "learning_rate": 8.228228228228229e-06,
+      "loss": 0.6216,
+      "step": 17700
+    },
+    {
+      "epoch": 53.18,
+      "grad_norm": 21.481840133666992,
+      "learning_rate": 8.227227227227229e-06,
+      "loss": 0.5999,
+      "step": 17710
+    },
+    {
+      "epoch": 53.21,
+      "grad_norm": 14.913869857788086,
+      "learning_rate": 8.226226226226227e-06,
+      "loss": 0.5996,
+      "step": 17720
+    },
+    {
+      "epoch": 53.24,
+      "grad_norm": 15.669997215270996,
+      "learning_rate": 8.225225225225226e-06,
+      "loss": 0.6066,
+      "step": 17730
+    },
+    {
+      "epoch": 53.27,
+      "grad_norm": 12.006368637084961,
+      "learning_rate": 8.224224224224225e-06,
+      "loss": 0.5889,
+      "step": 17740
+    },
+    {
+      "epoch": 53.3,
+      "grad_norm": 15.566818237304688,
+      "learning_rate": 8.223223223223223e-06,
+      "loss": 0.5884,
+      "step": 17750
+    },
+    {
+      "epoch": 53.33,
+      "grad_norm": 21.139230728149414,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 0.6272,
+      "step": 17760
+    },
+    {
+      "epoch": 53.36,
+      "grad_norm": 14.190361976623535,
+      "learning_rate": 8.221221221221222e-06,
+      "loss": 0.6257,
+      "step": 17770
+    },
+    {
+      "epoch": 53.39,
+      "grad_norm": 12.169196128845215,
+      "learning_rate": 8.22022022022022e-06,
+      "loss": 0.6332,
+      "step": 17780
+    },
+    {
+      "epoch": 53.42,
+      "grad_norm": 15.970104217529297,
+      "learning_rate": 8.21921921921922e-06,
+      "loss": 0.6271,
+      "step": 17790
+    },
+    {
+      "epoch": 53.45,
+      "grad_norm": 16.380125045776367,
+      "learning_rate": 8.21821821821822e-06,
+      "loss": 0.6672,
+      "step": 17800
+    },
+    {
+      "epoch": 53.48,
+      "grad_norm": 24.46523666381836,
+      "learning_rate": 8.217217217217218e-06,
+      "loss": 0.6545,
+      "step": 17810
+    },
+    {
+      "epoch": 53.51,
+      "grad_norm": 21.113473892211914,
+      "learning_rate": 8.216216216216217e-06,
+      "loss": 0.6436,
+      "step": 17820
+    },
+    {
+      "epoch": 53.54,
+      "grad_norm": 20.094146728515625,
+      "learning_rate": 8.215215215215215e-06,
+      "loss": 0.6114,
+      "step": 17830
+    },
+    {
+      "epoch": 53.57,
+      "grad_norm": 21.91627311706543,
+      "learning_rate": 8.214214214214214e-06,
+      "loss": 0.6301,
+      "step": 17840
+    },
+    {
+      "epoch": 53.6,
+      "grad_norm": 16.08312225341797,
+      "learning_rate": 8.213213213213214e-06,
+      "loss": 0.6304,
+      "step": 17850
+    },
+    {
+      "epoch": 53.63,
+      "grad_norm": 16.935136795043945,
+      "learning_rate": 8.212212212212213e-06,
+      "loss": 0.6028,
+      "step": 17860
+    },
+    {
+      "epoch": 53.66,
+      "grad_norm": 16.1990909576416,
+      "learning_rate": 8.211211211211211e-06,
+      "loss": 0.6504,
+      "step": 17870
+    },
+    {
+      "epoch": 53.69,
+      "grad_norm": 22.533565521240234,
+      "learning_rate": 8.210210210210212e-06,
+      "loss": 0.6307,
+      "step": 17880
+    },
+    {
+      "epoch": 53.72,
+      "grad_norm": 24.32695770263672,
+      "learning_rate": 8.20920920920921e-06,
+      "loss": 0.6518,
+      "step": 17890
+    },
+    {
+      "epoch": 53.75,
+      "grad_norm": 17.2038516998291,
+      "learning_rate": 8.208208208208209e-06,
+      "loss": 0.5854,
+      "step": 17900
+    },
+    {
+      "epoch": 53.78,
+      "grad_norm": 12.194356918334961,
+      "learning_rate": 8.20720720720721e-06,
+      "loss": 0.6106,
+      "step": 17910
+    },
+    {
+      "epoch": 53.81,
+      "grad_norm": 18.936695098876953,
+      "learning_rate": 8.206206206206206e-06,
+      "loss": 0.5942,
+      "step": 17920
+    },
+    {
+      "epoch": 53.84,
+      "grad_norm": 12.751850128173828,
+      "learning_rate": 8.205205205205206e-06,
+      "loss": 0.5833,
+      "step": 17930
+    },
+    {
+      "epoch": 53.87,
+      "grad_norm": 13.7149019241333,
+      "learning_rate": 8.204204204204205e-06,
+      "loss": 0.6136,
+      "step": 17940
+    },
+    {
+      "epoch": 53.9,
+      "grad_norm": 12.324130058288574,
+      "learning_rate": 8.203203203203204e-06,
+      "loss": 0.5969,
+      "step": 17950
+    },
+    {
+      "epoch": 53.93,
+      "grad_norm": 14.895933151245117,
+      "learning_rate": 8.202202202202202e-06,
+      "loss": 0.6523,
+      "step": 17960
+    },
+    {
+      "epoch": 53.96,
+      "grad_norm": 14.21236801147461,
+      "learning_rate": 8.201201201201202e-06,
+      "loss": 0.6516,
+      "step": 17970
+    },
+    {
+      "epoch": 53.99,
+      "grad_norm": 11.186808586120605,
+      "learning_rate": 8.200200200200201e-06,
+      "loss": 0.6018,
+      "step": 17980
+    },
+    {
+      "epoch": 54.0,
+      "eval_accuracy": 0.8668,
+      "eval_loss": 0.45088598132133484,
+      "eval_runtime": 13.0896,
+      "eval_samples_per_second": 763.966,
+      "eval_steps_per_second": 3.056,
+      "step": 17982
+    },
+    {
+      "epoch": 54.02,
+      "grad_norm": 11.544270515441895,
+      "learning_rate": 8.1991991991992e-06,
+      "loss": 0.5316,
+      "step": 17990
+    },
+    {
+      "epoch": 54.05,
+      "grad_norm": 18.69845199584961,
+      "learning_rate": 8.198198198198198e-06,
+      "loss": 0.6515,
+      "step": 18000
+    },
+    {
+      "epoch": 54.08,
+      "grad_norm": 14.782732009887695,
+      "learning_rate": 8.197197197197197e-06,
+      "loss": 0.658,
+      "step": 18010
+    },
+    {
+      "epoch": 54.11,
+      "grad_norm": 12.888195037841797,
+      "learning_rate": 8.196196196196197e-06,
+      "loss": 0.6079,
+      "step": 18020
+    },
+    {
+      "epoch": 54.14,
+      "grad_norm": 16.469144821166992,
+      "learning_rate": 8.195195195195196e-06,
+      "loss": 0.5935,
+      "step": 18030
+    },
+    {
+      "epoch": 54.17,
+      "grad_norm": 39.164031982421875,
+      "learning_rate": 8.194194194194194e-06,
+      "loss": 0.5693,
+      "step": 18040
+    },
+    {
+      "epoch": 54.2,
+      "grad_norm": 13.540485382080078,
+      "learning_rate": 8.193193193193195e-06,
+      "loss": 0.597,
+      "step": 18050
+    },
+    {
+      "epoch": 54.23,
+      "grad_norm": 18.07455062866211,
+      "learning_rate": 8.192192192192193e-06,
+      "loss": 0.6451,
+      "step": 18060
+    },
+    {
+      "epoch": 54.26,
+      "grad_norm": 14.38621997833252,
+      "learning_rate": 8.191191191191192e-06,
+      "loss": 0.5821,
+      "step": 18070
+    },
+    {
+      "epoch": 54.29,
+      "grad_norm": 18.677093505859375,
+      "learning_rate": 8.19019019019019e-06,
+      "loss": 0.622,
+      "step": 18080
+    },
+    {
+      "epoch": 54.32,
+      "grad_norm": 22.3408203125,
+      "learning_rate": 8.189189189189189e-06,
+      "loss": 0.6261,
+      "step": 18090
+    },
+    {
+      "epoch": 54.35,
+      "grad_norm": 11.709248542785645,
+      "learning_rate": 8.18818818818819e-06,
+      "loss": 0.6504,
+      "step": 18100
+    },
+    {
+      "epoch": 54.38,
+      "grad_norm": 15.678053855895996,
+      "learning_rate": 8.187187187187188e-06,
+      "loss": 0.6171,
+      "step": 18110
+    },
+    {
+      "epoch": 54.41,
+      "grad_norm": 13.658485412597656,
+      "learning_rate": 8.186186186186186e-06,
+      "loss": 0.562,
+      "step": 18120
+    },
+    {
+      "epoch": 54.44,
+      "grad_norm": 14.877076148986816,
+      "learning_rate": 8.185185185185187e-06,
+      "loss": 0.6325,
+      "step": 18130
+    },
+    {
+      "epoch": 54.47,
+      "grad_norm": 21.70797348022461,
+      "learning_rate": 8.184184184184185e-06,
+      "loss": 0.6752,
+      "step": 18140
+    },
+    {
+      "epoch": 54.5,
+      "grad_norm": 15.471793174743652,
+      "learning_rate": 8.183183183183184e-06,
+      "loss": 0.6115,
+      "step": 18150
+    },
+    {
+      "epoch": 54.53,
+      "grad_norm": 17.171215057373047,
+      "learning_rate": 8.182182182182183e-06,
+      "loss": 0.5683,
+      "step": 18160
+    },
+    {
+      "epoch": 54.56,
+      "grad_norm": 11.893485069274902,
+      "learning_rate": 8.181181181181181e-06,
+      "loss": 0.5831,
+      "step": 18170
+    },
+    {
+      "epoch": 54.59,
+      "grad_norm": 13.77318286895752,
+      "learning_rate": 8.18018018018018e-06,
+      "loss": 0.618,
+      "step": 18180
+    },
+    {
+      "epoch": 54.62,
+      "grad_norm": 12.6817045211792,
+      "learning_rate": 8.17917917917918e-06,
+      "loss": 0.5864,
+      "step": 18190
+    },
+    {
+      "epoch": 54.65,
+      "grad_norm": 12.999750137329102,
+      "learning_rate": 8.178178178178179e-06,
+      "loss": 0.5751,
+      "step": 18200
+    },
+    {
+      "epoch": 54.68,
+      "grad_norm": 11.285355567932129,
+      "learning_rate": 8.177177177177177e-06,
+      "loss": 0.6755,
+      "step": 18210
+    },
+    {
+      "epoch": 54.71,
+      "grad_norm": 24.20255470275879,
+      "learning_rate": 8.176176176176177e-06,
+      "loss": 0.621,
+      "step": 18220
+    },
+    {
+      "epoch": 54.74,
+      "grad_norm": 17.335365295410156,
+      "learning_rate": 8.175175175175176e-06,
+      "loss": 0.6793,
+      "step": 18230
+    },
+    {
+      "epoch": 54.77,
+      "grad_norm": 15.035238265991211,
+      "learning_rate": 8.174174174174175e-06,
+      "loss": 0.5939,
+      "step": 18240
+    },
+    {
+      "epoch": 54.8,
+      "grad_norm": 11.79697322845459,
+      "learning_rate": 8.173173173173173e-06,
+      "loss": 0.5754,
+      "step": 18250
+    },
+    {
+      "epoch": 54.83,
+      "grad_norm": 17.361902236938477,
+      "learning_rate": 8.172172172172172e-06,
+      "loss": 0.5622,
+      "step": 18260
+    },
+    {
+      "epoch": 54.86,
+      "grad_norm": 22.03786277770996,
+      "learning_rate": 8.171171171171172e-06,
+      "loss": 0.5917,
+      "step": 18270
+    },
+    {
+      "epoch": 54.89,
+      "grad_norm": 22.793529510498047,
+      "learning_rate": 8.17017017017017e-06,
+      "loss": 0.6358,
+      "step": 18280
+    },
+    {
+      "epoch": 54.92,
+      "grad_norm": 17.468017578125,
+      "learning_rate": 8.16916916916917e-06,
+      "loss": 0.607,
+      "step": 18290
+    },
+    {
+      "epoch": 54.95,
+      "grad_norm": 21.24018669128418,
+      "learning_rate": 8.16816816816817e-06,
+      "loss": 0.6604,
+      "step": 18300
+    },
+    {
+      "epoch": 54.98,
+      "grad_norm": 14.710421562194824,
+      "learning_rate": 8.167167167167168e-06,
+      "loss": 0.6068,
+      "step": 18310
+    },
+    {
+      "epoch": 55.0,
+      "eval_accuracy": 0.865,
+      "eval_loss": 0.45606616139411926,
+      "eval_runtime": 12.6186,
+      "eval_samples_per_second": 792.48,
+      "eval_steps_per_second": 3.17,
+      "step": 18315
+    },
+    {
+      "epoch": 55.02,
+      "grad_norm": 15.025124549865723,
+      "learning_rate": 8.166166166166167e-06,
+      "loss": 0.5491,
+      "step": 18320
+    },
+    {
+      "epoch": 55.05,
+      "grad_norm": 15.541799545288086,
+      "learning_rate": 8.165165165165165e-06,
+      "loss": 0.6588,
+      "step": 18330
+    },
+    {
+      "epoch": 55.08,
+      "grad_norm": 15.385310173034668,
+      "learning_rate": 8.164164164164164e-06,
+      "loss": 0.6377,
+      "step": 18340
+    },
+    {
+      "epoch": 55.11,
+      "grad_norm": 15.760037422180176,
+      "learning_rate": 8.163163163163164e-06,
+      "loss": 0.6153,
+      "step": 18350
+    },
+    {
+      "epoch": 55.14,
+      "grad_norm": 19.843534469604492,
+      "learning_rate": 8.162162162162163e-06,
+      "loss": 0.5795,
+      "step": 18360
+    },
+    {
+      "epoch": 55.17,
+      "grad_norm": 12.835470199584961,
+      "learning_rate": 8.161161161161161e-06,
+      "loss": 0.6025,
+      "step": 18370
+    },
+    {
+      "epoch": 55.2,
+      "grad_norm": 16.53463363647461,
+      "learning_rate": 8.160160160160162e-06,
+      "loss": 0.4885,
+      "step": 18380
+    },
+    {
+      "epoch": 55.23,
+      "grad_norm": 12.131781578063965,
+      "learning_rate": 8.15915915915916e-06,
+      "loss": 0.6007,
+      "step": 18390
+    },
+    {
+      "epoch": 55.26,
+      "grad_norm": 27.071044921875,
+      "learning_rate": 8.158158158158159e-06,
+      "loss": 0.5671,
+      "step": 18400
+    },
+    {
+      "epoch": 55.29,
+      "grad_norm": 17.2073974609375,
+      "learning_rate": 8.157157157157158e-06,
+      "loss": 0.5901,
+      "step": 18410
+    },
+    {
+      "epoch": 55.32,
+      "grad_norm": 15.685284614562988,
+      "learning_rate": 8.156156156156156e-06,
+      "loss": 0.5979,
+      "step": 18420
+    },
+    {
+      "epoch": 55.35,
+      "grad_norm": 11.316452980041504,
+      "learning_rate": 8.155155155155155e-06,
+      "loss": 0.5642,
+      "step": 18430
+    },
+    {
+      "epoch": 55.38,
+      "grad_norm": 15.418350219726562,
+      "learning_rate": 8.154154154154155e-06,
+      "loss": 0.6188,
+      "step": 18440
+    },
+    {
+      "epoch": 55.41,
+      "grad_norm": 15.311288833618164,
+      "learning_rate": 8.153153153153154e-06,
+      "loss": 0.6231,
+      "step": 18450
+    },
+    {
+      "epoch": 55.44,
+      "grad_norm": 18.07636260986328,
+      "learning_rate": 8.152152152152152e-06,
+      "loss": 0.6114,
+      "step": 18460
+    },
+    {
+      "epoch": 55.47,
+      "grad_norm": 15.564066886901855,
+      "learning_rate": 8.151151151151153e-06,
+      "loss": 0.5485,
+      "step": 18470
+    },
+    {
+      "epoch": 55.5,
+      "grad_norm": 12.460175514221191,
+      "learning_rate": 8.150150150150151e-06,
+      "loss": 0.5263,
+      "step": 18480
+    },
+    {
+      "epoch": 55.53,
+      "grad_norm": 18.801799774169922,
+      "learning_rate": 8.14914914914915e-06,
+      "loss": 0.6383,
+      "step": 18490
+    },
+    {
+      "epoch": 55.56,
+      "grad_norm": 14.195170402526855,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.5956,
+      "step": 18500
+    },
+    {
+      "epoch": 55.59,
+      "grad_norm": 16.936365127563477,
+      "learning_rate": 8.147147147147147e-06,
+      "loss": 0.5915,
+      "step": 18510
+    },
+    {
+      "epoch": 55.62,
+      "grad_norm": 11.47391414642334,
+      "learning_rate": 8.146146146146147e-06,
+      "loss": 0.5321,
+      "step": 18520
+    },
+    {
+      "epoch": 55.65,
+      "grad_norm": 18.33390235900879,
+      "learning_rate": 8.145145145145146e-06,
+      "loss": 0.6296,
+      "step": 18530
+    },
+    {
+      "epoch": 55.68,
+      "grad_norm": 27.2676944732666,
+      "learning_rate": 8.144144144144144e-06,
+      "loss": 0.5628,
+      "step": 18540
+    },
+    {
+      "epoch": 55.71,
+      "grad_norm": 16.5135555267334,
+      "learning_rate": 8.143143143143145e-06,
+      "loss": 0.5631,
+      "step": 18550
+    },
+    {
+      "epoch": 55.74,
+      "grad_norm": 20.772871017456055,
+      "learning_rate": 8.142142142142143e-06,
+      "loss": 0.6028,
+      "step": 18560
+    },
+    {
+      "epoch": 55.77,
+      "grad_norm": 18.499374389648438,
+      "learning_rate": 8.141141141141142e-06,
+      "loss": 0.6235,
+      "step": 18570
+    },
+    {
+      "epoch": 55.8,
+      "grad_norm": 20.60066032409668,
+      "learning_rate": 8.14014014014014e-06,
+      "loss": 0.5681,
+      "step": 18580
+    },
+    {
+      "epoch": 55.83,
+      "grad_norm": 16.534555435180664,
+      "learning_rate": 8.139139139139139e-06,
+      "loss": 0.6002,
+      "step": 18590
+    },
+    {
+      "epoch": 55.86,
+      "grad_norm": 20.2315673828125,
+      "learning_rate": 8.13813813813814e-06,
+      "loss": 0.5633,
+      "step": 18600
+    },
+    {
+      "epoch": 55.89,
+      "grad_norm": 18.77223014831543,
+      "learning_rate": 8.137137137137138e-06,
+      "loss": 0.6241,
+      "step": 18610
+    },
+    {
+      "epoch": 55.92,
+      "grad_norm": 16.95858383178711,
+      "learning_rate": 8.136136136136137e-06,
+      "loss": 0.5785,
+      "step": 18620
+    },
+    {
+      "epoch": 55.95,
+      "grad_norm": 22.44672393798828,
+      "learning_rate": 8.135135135135137e-06,
+      "loss": 0.6005,
+      "step": 18630
+    },
+    {
+      "epoch": 55.98,
+      "grad_norm": 12.324645042419434,
+      "learning_rate": 8.134134134134135e-06,
+      "loss": 0.6368,
+      "step": 18640
+    },
+    {
+      "epoch": 56.0,
+      "eval_accuracy": 0.8666,
+      "eval_loss": 0.4533466696739197,
+      "eval_runtime": 12.8866,
+      "eval_samples_per_second": 776.001,
+      "eval_steps_per_second": 3.104,
+      "step": 18648
+    },
+    {
+      "epoch": 56.01,
+      "grad_norm": 18.5239200592041,
+      "learning_rate": 8.133133133133134e-06,
+      "loss": 0.6064,
+      "step": 18650
+    },
+    {
+      "epoch": 56.04,
+      "grad_norm": 12.947190284729004,
+      "learning_rate": 8.132132132132133e-06,
+      "loss": 0.5789,
+      "step": 18660
+    },
+    {
+      "epoch": 56.07,
+      "grad_norm": 15.244621276855469,
+      "learning_rate": 8.131131131131131e-06,
+      "loss": 0.5906,
+      "step": 18670
+    },
+    {
+      "epoch": 56.1,
+      "grad_norm": 19.841922760009766,
+      "learning_rate": 8.13013013013013e-06,
+      "loss": 0.5527,
+      "step": 18680
+    },
+    {
+      "epoch": 56.13,
+      "grad_norm": 16.695791244506836,
+      "learning_rate": 8.12912912912913e-06,
+      "loss": 0.5803,
+      "step": 18690
+    },
+    {
+      "epoch": 56.16,
+      "grad_norm": 19.614633560180664,
+      "learning_rate": 8.128128128128129e-06,
+      "loss": 0.6265,
+      "step": 18700
+    },
+    {
+      "epoch": 56.19,
+      "grad_norm": 11.918386459350586,
+      "learning_rate": 8.127127127127127e-06,
+      "loss": 0.5851,
+      "step": 18710
+    },
+    {
+      "epoch": 56.22,
+      "grad_norm": 12.749957084655762,
+      "learning_rate": 8.126126126126128e-06,
+      "loss": 0.6859,
+      "step": 18720
+    },
+    {
+      "epoch": 56.25,
+      "grad_norm": 17.98277473449707,
+      "learning_rate": 8.125125125125126e-06,
+      "loss": 0.5873,
+      "step": 18730
+    },
+    {
+      "epoch": 56.28,
+      "grad_norm": 25.136402130126953,
+      "learning_rate": 8.124124124124125e-06,
+      "loss": 0.6042,
+      "step": 18740
+    },
+    {
+      "epoch": 56.31,
+      "grad_norm": 17.568328857421875,
+      "learning_rate": 8.123123123123123e-06,
+      "loss": 0.5497,
+      "step": 18750
+    },
+    {
+      "epoch": 56.34,
+      "grad_norm": 16.162120819091797,
+      "learning_rate": 8.122122122122122e-06,
+      "loss": 0.5819,
+      "step": 18760
+    },
+    {
+      "epoch": 56.37,
+      "grad_norm": 15.463545799255371,
+      "learning_rate": 8.121121121121122e-06,
+      "loss": 0.5787,
+      "step": 18770
+    },
+    {
+      "epoch": 56.4,
+      "grad_norm": 14.494797706604004,
+      "learning_rate": 8.12012012012012e-06,
+      "loss": 0.5883,
+      "step": 18780
+    },
+    {
+      "epoch": 56.43,
+      "grad_norm": 16.26420021057129,
+      "learning_rate": 8.11911911911912e-06,
+      "loss": 0.5825,
+      "step": 18790
+    },
+    {
+      "epoch": 56.46,
+      "grad_norm": 16.86106300354004,
+      "learning_rate": 8.11811811811812e-06,
+      "loss": 0.5885,
+      "step": 18800
+    },
+    {
+      "epoch": 56.49,
+      "grad_norm": 17.785783767700195,
+      "learning_rate": 8.117117117117118e-06,
+      "loss": 0.5989,
+      "step": 18810
+    },
+    {
+      "epoch": 56.52,
+      "grad_norm": 21.74614906311035,
+      "learning_rate": 8.116116116116117e-06,
+      "loss": 0.6554,
+      "step": 18820
+    },
+    {
+      "epoch": 56.55,
+      "grad_norm": 15.670501708984375,
+      "learning_rate": 8.115115115115115e-06,
+      "loss": 0.5839,
+      "step": 18830
+    },
+    {
+      "epoch": 56.58,
+      "grad_norm": 14.93682861328125,
+      "learning_rate": 8.114114114114114e-06,
+      "loss": 0.5679,
+      "step": 18840
+    },
+    {
+      "epoch": 56.61,
+      "grad_norm": 12.969562530517578,
+      "learning_rate": 8.113113113113114e-06,
+      "loss": 0.6219,
+      "step": 18850
+    },
+    {
+      "epoch": 56.64,
+      "grad_norm": 13.447518348693848,
+      "learning_rate": 8.112112112112113e-06,
+      "loss": 0.579,
+      "step": 18860
+    },
+    {
+      "epoch": 56.67,
+      "grad_norm": 12.603790283203125,
+      "learning_rate": 8.111111111111112e-06,
+      "loss": 0.6207,
+      "step": 18870
+    },
+    {
+      "epoch": 56.7,
+      "grad_norm": 13.396700859069824,
+      "learning_rate": 8.11011011011011e-06,
+      "loss": 0.5672,
+      "step": 18880
+    },
+    {
+      "epoch": 56.73,
+      "grad_norm": 22.983783721923828,
+      "learning_rate": 8.10910910910911e-06,
+      "loss": 0.6262,
+      "step": 18890
+    },
+    {
+      "epoch": 56.76,
+      "grad_norm": 22.194602966308594,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.5478,
+      "step": 18900
+    },
+    {
+      "epoch": 56.79,
+      "grad_norm": 14.818936347961426,
+      "learning_rate": 8.107107107107108e-06,
+      "loss": 0.6087,
+      "step": 18910
+    },
+    {
+      "epoch": 56.82,
+      "grad_norm": 14.03148365020752,
+      "learning_rate": 8.106106106106106e-06,
+      "loss": 0.5593,
+      "step": 18920
+    },
+    {
+      "epoch": 56.85,
+      "grad_norm": 13.037586212158203,
+      "learning_rate": 8.105105105105105e-06,
+      "loss": 0.5773,
+      "step": 18930
+    },
+    {
+      "epoch": 56.88,
+      "grad_norm": 19.797903060913086,
+      "learning_rate": 8.104104104104105e-06,
+      "loss": 0.6012,
+      "step": 18940
+    },
+    {
+      "epoch": 56.91,
+      "grad_norm": 12.080364227294922,
+      "learning_rate": 8.103103103103104e-06,
+      "loss": 0.6236,
+      "step": 18950
+    },
+    {
+      "epoch": 56.94,
+      "grad_norm": 14.649006843566895,
+      "learning_rate": 8.102102102102102e-06,
+      "loss": 0.646,
+      "step": 18960
+    },
+    {
+      "epoch": 56.97,
+      "grad_norm": 15.998294830322266,
+      "learning_rate": 8.101101101101103e-06,
+      "loss": 0.6381,
+      "step": 18970
+    },
+    {
+      "epoch": 57.0,
+      "grad_norm": 15.81445598602295,
+      "learning_rate": 8.100100100100101e-06,
+      "loss": 0.5945,
+      "step": 18980
+    },
+    {
+      "epoch": 57.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.45373043417930603,
+      "eval_runtime": 13.1449,
+      "eval_samples_per_second": 760.752,
+      "eval_steps_per_second": 3.043,
+      "step": 18981
+    },
+    {
+      "epoch": 57.03,
+      "grad_norm": 10.915128707885742,
+      "learning_rate": 8.0990990990991e-06,
+      "loss": 0.6811,
+      "step": 18990
+    },
+    {
+      "epoch": 57.06,
+      "grad_norm": 16.743343353271484,
+      "learning_rate": 8.098098098098098e-06,
+      "loss": 0.6148,
+      "step": 19000
+    },
+    {
+      "epoch": 57.09,
+      "grad_norm": 11.98512077331543,
+      "learning_rate": 8.097097097097097e-06,
+      "loss": 0.5817,
+      "step": 19010
+    },
+    {
+      "epoch": 57.12,
+      "grad_norm": 12.866323471069336,
+      "learning_rate": 8.096096096096097e-06,
+      "loss": 0.5966,
+      "step": 19020
+    },
+    {
+      "epoch": 57.15,
+      "grad_norm": 13.117215156555176,
+      "learning_rate": 8.095095095095096e-06,
+      "loss": 0.5596,
+      "step": 19030
+    },
+    {
+      "epoch": 57.18,
+      "grad_norm": 22.87441635131836,
+      "learning_rate": 8.094094094094094e-06,
+      "loss": 0.5684,
+      "step": 19040
+    },
+    {
+      "epoch": 57.21,
+      "grad_norm": 10.847088813781738,
+      "learning_rate": 8.093093093093095e-06,
+      "loss": 0.5343,
+      "step": 19050
+    },
+    {
+      "epoch": 57.24,
+      "grad_norm": 18.142459869384766,
+      "learning_rate": 8.092092092092093e-06,
+      "loss": 0.5796,
+      "step": 19060
+    },
+    {
+      "epoch": 57.27,
+      "grad_norm": 19.32158660888672,
+      "learning_rate": 8.091091091091092e-06,
+      "loss": 0.5566,
+      "step": 19070
+    },
+    {
+      "epoch": 57.3,
+      "grad_norm": 13.15195369720459,
+      "learning_rate": 8.09009009009009e-06,
+      "loss": 0.5322,
+      "step": 19080
+    },
+    {
+      "epoch": 57.33,
+      "grad_norm": 14.730746269226074,
+      "learning_rate": 8.089089089089089e-06,
+      "loss": 0.6219,
+      "step": 19090
+    },
+    {
+      "epoch": 57.36,
+      "grad_norm": 16.131155014038086,
+      "learning_rate": 8.088088088088088e-06,
+      "loss": 0.5773,
+      "step": 19100
+    },
+    {
+      "epoch": 57.39,
+      "grad_norm": 18.815977096557617,
+      "learning_rate": 8.087087087087088e-06,
+      "loss": 0.5456,
+      "step": 19110
+    },
+    {
+      "epoch": 57.42,
+      "grad_norm": 16.685556411743164,
+      "learning_rate": 8.086086086086087e-06,
+      "loss": 0.5711,
+      "step": 19120
+    },
+    {
+      "epoch": 57.45,
+      "grad_norm": 17.469432830810547,
+      "learning_rate": 8.085085085085085e-06,
+      "loss": 0.5864,
+      "step": 19130
+    },
+    {
+      "epoch": 57.48,
+      "grad_norm": 17.60506248474121,
+      "learning_rate": 8.084084084084085e-06,
+      "loss": 0.6233,
+      "step": 19140
+    },
+    {
+      "epoch": 57.51,
+      "grad_norm": 14.348414421081543,
+      "learning_rate": 8.083083083083084e-06,
+      "loss": 0.5357,
+      "step": 19150
+    },
+    {
+      "epoch": 57.54,
+      "grad_norm": 19.012470245361328,
+      "learning_rate": 8.082082082082083e-06,
+      "loss": 0.6338,
+      "step": 19160
+    },
+    {
+      "epoch": 57.57,
+      "grad_norm": 13.458383560180664,
+      "learning_rate": 8.081081081081081e-06,
+      "loss": 0.5687,
+      "step": 19170
+    },
+    {
+      "epoch": 57.6,
+      "grad_norm": 18.919784545898438,
+      "learning_rate": 8.08008008008008e-06,
+      "loss": 0.6034,
+      "step": 19180
+    },
+    {
+      "epoch": 57.63,
+      "grad_norm": 16.603261947631836,
+      "learning_rate": 8.07907907907908e-06,
+      "loss": 0.6172,
+      "step": 19190
+    },
+    {
+      "epoch": 57.66,
+      "grad_norm": 11.052780151367188,
+      "learning_rate": 8.078078078078079e-06,
+      "loss": 0.6061,
+      "step": 19200
+    },
+    {
+      "epoch": 57.69,
+      "grad_norm": 12.503372192382812,
+      "learning_rate": 8.077077077077077e-06,
+      "loss": 0.6252,
+      "step": 19210
+    },
+    {
+      "epoch": 57.72,
+      "grad_norm": 14.037562370300293,
+      "learning_rate": 8.076076076076078e-06,
+      "loss": 0.6156,
+      "step": 19220
+    },
+    {
+      "epoch": 57.75,
+      "grad_norm": 20.213315963745117,
+      "learning_rate": 8.075075075075076e-06,
+      "loss": 0.6186,
+      "step": 19230
+    },
+    {
+      "epoch": 57.78,
+      "grad_norm": 18.24369239807129,
+      "learning_rate": 8.074074074074075e-06,
+      "loss": 0.5665,
+      "step": 19240
+    },
+    {
+      "epoch": 57.81,
+      "grad_norm": 15.676369667053223,
+      "learning_rate": 8.073073073073073e-06,
+      "loss": 0.6401,
+      "step": 19250
+    },
+    {
+      "epoch": 57.84,
+      "grad_norm": 19.942174911499023,
+      "learning_rate": 8.072072072072072e-06,
+      "loss": 0.5869,
+      "step": 19260
+    },
+    {
+      "epoch": 57.87,
+      "grad_norm": 18.580678939819336,
+      "learning_rate": 8.071071071071072e-06,
+      "loss": 0.594,
+      "step": 19270
+    },
+    {
+      "epoch": 57.9,
+      "grad_norm": 14.152729988098145,
+      "learning_rate": 8.070070070070071e-06,
+      "loss": 0.6364,
+      "step": 19280
+    },
+    {
+      "epoch": 57.93,
+      "grad_norm": 15.426253318786621,
+      "learning_rate": 8.06906906906907e-06,
+      "loss": 0.6143,
+      "step": 19290
+    },
+    {
+      "epoch": 57.96,
+      "grad_norm": 12.899490356445312,
+      "learning_rate": 8.06806806806807e-06,
+      "loss": 0.567,
+      "step": 19300
+    },
+    {
+      "epoch": 57.99,
+      "grad_norm": 13.860706329345703,
+      "learning_rate": 8.067067067067068e-06,
+      "loss": 0.5379,
+      "step": 19310
+    },
+    {
+      "epoch": 58.0,
+      "eval_accuracy": 0.8644,
+      "eval_loss": 0.4582826793193817,
+      "eval_runtime": 13.2356,
+      "eval_samples_per_second": 755.537,
+      "eval_steps_per_second": 3.022,
+      "step": 19314
+    },
+    {
+      "epoch": 58.02,
+      "grad_norm": 15.111919403076172,
+      "learning_rate": 8.066066066066067e-06,
+      "loss": 0.544,
+      "step": 19320
+    },
+    {
+      "epoch": 58.05,
+      "grad_norm": 15.057886123657227,
+      "learning_rate": 8.065065065065066e-06,
+      "loss": 0.5784,
+      "step": 19330
+    },
+    {
+      "epoch": 58.08,
+      "grad_norm": 19.320295333862305,
+      "learning_rate": 8.064064064064064e-06,
+      "loss": 0.6362,
+      "step": 19340
+    },
+    {
+      "epoch": 58.11,
+      "grad_norm": 18.68480110168457,
+      "learning_rate": 8.063063063063063e-06,
+      "loss": 0.6141,
+      "step": 19350
+    },
+    {
+      "epoch": 58.14,
+      "grad_norm": 16.06761360168457,
+      "learning_rate": 8.062062062062063e-06,
+      "loss": 0.5976,
+      "step": 19360
+    },
+    {
+      "epoch": 58.17,
+      "grad_norm": 32.325294494628906,
+      "learning_rate": 8.061061061061062e-06,
+      "loss": 0.5454,
+      "step": 19370
+    },
+    {
+      "epoch": 58.2,
+      "grad_norm": 19.5351619720459,
+      "learning_rate": 8.06006006006006e-06,
+      "loss": 0.6339,
+      "step": 19380
+    },
+    {
+      "epoch": 58.23,
+      "grad_norm": 22.733434677124023,
+      "learning_rate": 8.05905905905906e-06,
+      "loss": 0.6224,
+      "step": 19390
+    },
+    {
+      "epoch": 58.26,
+      "grad_norm": 16.71336555480957,
+      "learning_rate": 8.058058058058059e-06,
+      "loss": 0.6396,
+      "step": 19400
+    },
+    {
+      "epoch": 58.29,
+      "grad_norm": 13.782090187072754,
+      "learning_rate": 8.057057057057058e-06,
+      "loss": 0.5882,
+      "step": 19410
+    },
+    {
+      "epoch": 58.32,
+      "grad_norm": 15.715386390686035,
+      "learning_rate": 8.056056056056056e-06,
+      "loss": 0.5647,
+      "step": 19420
+    },
+    {
+      "epoch": 58.35,
+      "grad_norm": 17.997074127197266,
+      "learning_rate": 8.055055055055055e-06,
+      "loss": 0.6363,
+      "step": 19430
+    },
+    {
+      "epoch": 58.38,
+      "grad_norm": 14.587823867797852,
+      "learning_rate": 8.054054054054055e-06,
+      "loss": 0.5845,
+      "step": 19440
+    },
+    {
+      "epoch": 58.41,
+      "grad_norm": 15.96159839630127,
+      "learning_rate": 8.053053053053054e-06,
+      "loss": 0.5909,
+      "step": 19450
+    },
+    {
+      "epoch": 58.44,
+      "grad_norm": 12.311030387878418,
+      "learning_rate": 8.052052052052052e-06,
+      "loss": 0.5711,
+      "step": 19460
+    },
+    {
+      "epoch": 58.47,
+      "grad_norm": 16.868986129760742,
+      "learning_rate": 8.051051051051053e-06,
+      "loss": 0.5818,
+      "step": 19470
+    },
+    {
+      "epoch": 58.5,
+      "grad_norm": 18.937164306640625,
+      "learning_rate": 8.050050050050051e-06,
+      "loss": 0.6229,
+      "step": 19480
+    },
+    {
+      "epoch": 58.53,
+      "grad_norm": 12.239059448242188,
+      "learning_rate": 8.04904904904905e-06,
+      "loss": 0.5801,
+      "step": 19490
+    },
+    {
+      "epoch": 58.56,
+      "grad_norm": 15.623565673828125,
+      "learning_rate": 8.048048048048048e-06,
+      "loss": 0.6241,
+      "step": 19500
+    },
+    {
+      "epoch": 58.59,
+      "grad_norm": 21.60228157043457,
+      "learning_rate": 8.047047047047047e-06,
+      "loss": 0.5768,
+      "step": 19510
+    },
+    {
+      "epoch": 58.62,
+      "grad_norm": 13.131705284118652,
+      "learning_rate": 8.046046046046047e-06,
+      "loss": 0.5423,
+      "step": 19520
+    },
+    {
+      "epoch": 58.65,
+      "grad_norm": 16.483760833740234,
+      "learning_rate": 8.045045045045046e-06,
+      "loss": 0.5831,
+      "step": 19530
+    },
+    {
+      "epoch": 58.68,
+      "grad_norm": 14.74811840057373,
+      "learning_rate": 8.044044044044045e-06,
+      "loss": 0.5568,
+      "step": 19540
+    },
+    {
+      "epoch": 58.71,
+      "grad_norm": 18.404155731201172,
+      "learning_rate": 8.043043043043043e-06,
+      "loss": 0.5936,
+      "step": 19550
+    },
+    {
+      "epoch": 58.74,
+      "grad_norm": 14.674580574035645,
+      "learning_rate": 8.042042042042043e-06,
+      "loss": 0.5879,
+      "step": 19560
+    },
+    {
+      "epoch": 58.77,
+      "grad_norm": 23.461597442626953,
+      "learning_rate": 8.041041041041042e-06,
+      "loss": 0.5931,
+      "step": 19570
+    },
+    {
+      "epoch": 58.8,
+      "grad_norm": 24.857013702392578,
+      "learning_rate": 8.04004004004004e-06,
+      "loss": 0.6067,
+      "step": 19580
+    },
+    {
+      "epoch": 58.83,
+      "grad_norm": 16.145484924316406,
+      "learning_rate": 8.03903903903904e-06,
+      "loss": 0.6382,
+      "step": 19590
+    },
+    {
+      "epoch": 58.86,
+      "grad_norm": 19.54372787475586,
+      "learning_rate": 8.038038038038038e-06,
+      "loss": 0.6105,
+      "step": 19600
+    },
+    {
+      "epoch": 58.89,
+      "grad_norm": 12.304158210754395,
+      "learning_rate": 8.037037037037038e-06,
+      "loss": 0.5826,
+      "step": 19610
+    },
+    {
+      "epoch": 58.92,
+      "grad_norm": 13.34826374053955,
+      "learning_rate": 8.036036036036037e-06,
+      "loss": 0.5916,
+      "step": 19620
+    },
+    {
+      "epoch": 58.95,
+      "grad_norm": 15.727002143859863,
+      "learning_rate": 8.035035035035035e-06,
+      "loss": 0.5416,
+      "step": 19630
+    },
+    {
+      "epoch": 58.98,
+      "grad_norm": 20.862152099609375,
+      "learning_rate": 8.034034034034036e-06,
+      "loss": 0.6031,
+      "step": 19640
+    },
+    {
+      "epoch": 59.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.4573589861392975,
+      "eval_runtime": 12.7266,
+      "eval_samples_per_second": 785.754,
+      "eval_steps_per_second": 3.143,
+      "step": 19647
+    },
+    {
+      "epoch": 59.01,
+      "grad_norm": 17.41853904724121,
+      "learning_rate": 8.033033033033034e-06,
+      "loss": 0.5721,
+      "step": 19650
+    },
+    {
+      "epoch": 59.04,
+      "grad_norm": 14.630786895751953,
+      "learning_rate": 8.032032032032033e-06,
+      "loss": 0.5814,
+      "step": 19660
+    },
+    {
+      "epoch": 59.07,
+      "grad_norm": 18.51511573791504,
+      "learning_rate": 8.031031031031031e-06,
+      "loss": 0.6257,
+      "step": 19670
+    },
+    {
+      "epoch": 59.1,
+      "grad_norm": 16.176118850708008,
+      "learning_rate": 8.03003003003003e-06,
+      "loss": 0.5536,
+      "step": 19680
+    },
+    {
+      "epoch": 59.13,
+      "grad_norm": 18.753887176513672,
+      "learning_rate": 8.02902902902903e-06,
+      "loss": 0.5978,
+      "step": 19690
+    },
+    {
+      "epoch": 59.16,
+      "grad_norm": 14.25525188446045,
+      "learning_rate": 8.028028028028029e-06,
+      "loss": 0.535,
+      "step": 19700
+    },
+    {
+      "epoch": 59.19,
+      "grad_norm": 23.966169357299805,
+      "learning_rate": 8.027027027027027e-06,
+      "loss": 0.5691,
+      "step": 19710
+    },
+    {
+      "epoch": 59.22,
+      "grad_norm": 15.448729515075684,
+      "learning_rate": 8.026026026026028e-06,
+      "loss": 0.6087,
+      "step": 19720
+    },
+    {
+      "epoch": 59.25,
+      "grad_norm": 14.6622314453125,
+      "learning_rate": 8.025025025025026e-06,
+      "loss": 0.5718,
+      "step": 19730
+    },
+    {
+      "epoch": 59.28,
+      "grad_norm": 28.261085510253906,
+      "learning_rate": 8.024024024024025e-06,
+      "loss": 0.5416,
+      "step": 19740
+    },
+    {
+      "epoch": 59.31,
+      "grad_norm": 19.39133071899414,
+      "learning_rate": 8.023023023023023e-06,
+      "loss": 0.5686,
+      "step": 19750
+    },
+    {
+      "epoch": 59.34,
+      "grad_norm": 14.506977081298828,
+      "learning_rate": 8.022022022022022e-06,
+      "loss": 0.5647,
+      "step": 19760
+    },
+    {
+      "epoch": 59.37,
+      "grad_norm": 11.523022651672363,
+      "learning_rate": 8.021021021021022e-06,
+      "loss": 0.5916,
+      "step": 19770
+    },
+    {
+      "epoch": 59.4,
+      "grad_norm": 14.599289894104004,
+      "learning_rate": 8.020020020020021e-06,
+      "loss": 0.6119,
+      "step": 19780
+    },
+    {
+      "epoch": 59.43,
+      "grad_norm": 16.842275619506836,
+      "learning_rate": 8.01901901901902e-06,
+      "loss": 0.6193,
+      "step": 19790
+    },
+    {
+      "epoch": 59.46,
+      "grad_norm": 17.81415367126465,
+      "learning_rate": 8.018018018018018e-06,
+      "loss": 0.653,
+      "step": 19800
+    },
+    {
+      "epoch": 59.49,
+      "grad_norm": 13.608844757080078,
+      "learning_rate": 8.017017017017018e-06,
+      "loss": 0.6132,
+      "step": 19810
+    },
+    {
+      "epoch": 59.52,
+      "grad_norm": 17.850231170654297,
+      "learning_rate": 8.016016016016017e-06,
+      "loss": 0.5948,
+      "step": 19820
+    },
+    {
+      "epoch": 59.55,
+      "grad_norm": 14.544975280761719,
+      "learning_rate": 8.015015015015016e-06,
+      "loss": 0.5357,
+      "step": 19830
+    },
+    {
+      "epoch": 59.58,
+      "grad_norm": 22.349634170532227,
+      "learning_rate": 8.014014014014014e-06,
+      "loss": 0.5214,
+      "step": 19840
+    },
+    {
+      "epoch": 59.61,
+      "grad_norm": 14.013883590698242,
+      "learning_rate": 8.013013013013013e-06,
+      "loss": 0.496,
+      "step": 19850
+    },
+    {
+      "epoch": 59.64,
+      "grad_norm": 13.427343368530273,
+      "learning_rate": 8.012012012012013e-06,
+      "loss": 0.56,
+      "step": 19860
+    },
+    {
+      "epoch": 59.67,
+      "grad_norm": 14.705164909362793,
+      "learning_rate": 8.011011011011012e-06,
+      "loss": 0.5356,
+      "step": 19870
+    },
+    {
+      "epoch": 59.7,
+      "grad_norm": 20.24700164794922,
+      "learning_rate": 8.01001001001001e-06,
+      "loss": 0.6188,
+      "step": 19880
+    },
+    {
+      "epoch": 59.73,
+      "grad_norm": 16.095613479614258,
+      "learning_rate": 8.00900900900901e-06,
+      "loss": 0.6366,
+      "step": 19890
+    },
+    {
+      "epoch": 59.76,
+      "grad_norm": 14.603818893432617,
+      "learning_rate": 8.00800800800801e-06,
+      "loss": 0.5962,
+      "step": 19900
+    },
+    {
+      "epoch": 59.79,
+      "grad_norm": 25.40558624267578,
+      "learning_rate": 8.007007007007008e-06,
+      "loss": 0.6345,
+      "step": 19910
+    },
+    {
+      "epoch": 59.82,
+      "grad_norm": 15.939118385314941,
+      "learning_rate": 8.006006006006006e-06,
+      "loss": 0.5625,
+      "step": 19920
+    },
+    {
+      "epoch": 59.85,
+      "grad_norm": 24.41084098815918,
+      "learning_rate": 8.005005005005005e-06,
+      "loss": 0.6134,
+      "step": 19930
+    },
+    {
+      "epoch": 59.88,
+      "grad_norm": 12.29263687133789,
+      "learning_rate": 8.004004004004005e-06,
+      "loss": 0.5302,
+      "step": 19940
+    },
+    {
+      "epoch": 59.91,
+      "grad_norm": 17.57611846923828,
+      "learning_rate": 8.003003003003004e-06,
+      "loss": 0.587,
+      "step": 19950
+    },
+    {
+      "epoch": 59.94,
+      "grad_norm": 12.565730094909668,
+      "learning_rate": 8.002002002002002e-06,
+      "loss": 0.5219,
+      "step": 19960
+    },
+    {
+      "epoch": 59.97,
+      "grad_norm": 20.622846603393555,
+      "learning_rate": 8.001001001001003e-06,
+      "loss": 0.6452,
+      "step": 19970
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 131.89938354492188,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.5445,
+      "step": 19980
+    },
+    {
+      "epoch": 60.0,
+      "eval_accuracy": 0.8629,
+      "eval_loss": 0.4607356786727905,
+      "eval_runtime": 13.0912,
+      "eval_samples_per_second": 763.874,
+      "eval_steps_per_second": 3.055,
+      "step": 19980
+    },
+    {
+      "epoch": 60.03,
+      "grad_norm": 14.892760276794434,
+      "learning_rate": 7.998998998999e-06,
+      "loss": 0.5442,
+      "step": 19990
+    },
+    {
+      "epoch": 60.06,
+      "grad_norm": 15.569472312927246,
+      "learning_rate": 7.997997997997999e-06,
+      "loss": 0.6092,
+      "step": 20000
+    },
+    {
+      "epoch": 60.09,
+      "grad_norm": 18.588777542114258,
+      "learning_rate": 7.996996996996997e-06,
+      "loss": 0.5771,
+      "step": 20010
+    },
+    {
+      "epoch": 60.12,
+      "grad_norm": 21.79499626159668,
+      "learning_rate": 7.995995995995996e-06,
+      "loss": 0.6436,
+      "step": 20020
+    },
+    {
+      "epoch": 60.15,
+      "grad_norm": 10.009160041809082,
+      "learning_rate": 7.994994994994996e-06,
+      "loss": 0.5526,
+      "step": 20030
+    },
+    {
+      "epoch": 60.18,
+      "grad_norm": 14.873530387878418,
+      "learning_rate": 7.993993993993995e-06,
+      "loss": 0.5506,
+      "step": 20040
+    },
+    {
+      "epoch": 60.21,
+      "grad_norm": 14.862759590148926,
+      "learning_rate": 7.992992992992993e-06,
+      "loss": 0.5676,
+      "step": 20050
+    },
+    {
+      "epoch": 60.24,
+      "grad_norm": 14.91917896270752,
+      "learning_rate": 7.991991991991993e-06,
+      "loss": 0.5795,
+      "step": 20060
+    },
+    {
+      "epoch": 60.27,
+      "grad_norm": 16.91362762451172,
+      "learning_rate": 7.990990990990992e-06,
+      "loss": 0.5699,
+      "step": 20070
+    },
+    {
+      "epoch": 60.3,
+      "grad_norm": 22.361339569091797,
+      "learning_rate": 7.98998998998999e-06,
+      "loss": 0.5749,
+      "step": 20080
+    },
+    {
+      "epoch": 60.33,
+      "grad_norm": 25.052778244018555,
+      "learning_rate": 7.98898898898899e-06,
+      "loss": 0.5692,
+      "step": 20090
+    },
+    {
+      "epoch": 60.36,
+      "grad_norm": 19.123653411865234,
+      "learning_rate": 7.987987987987988e-06,
+      "loss": 0.5832,
+      "step": 20100
+    },
+    {
+      "epoch": 60.39,
+      "grad_norm": 16.62225341796875,
+      "learning_rate": 7.986986986986988e-06,
+      "loss": 0.5846,
+      "step": 20110
+    },
+    {
+      "epoch": 60.42,
+      "grad_norm": 16.223085403442383,
+      "learning_rate": 7.985985985985987e-06,
+      "loss": 0.5737,
+      "step": 20120
+    },
+    {
+      "epoch": 60.45,
+      "grad_norm": 20.05699920654297,
+      "learning_rate": 7.984984984984985e-06,
+      "loss": 0.5762,
+      "step": 20130
+    },
+    {
+      "epoch": 60.48,
+      "grad_norm": 16.24242401123047,
+      "learning_rate": 7.983983983983986e-06,
+      "loss": 0.553,
+      "step": 20140
+    },
+    {
+      "epoch": 60.51,
+      "grad_norm": 12.968668937683105,
+      "learning_rate": 7.982982982982984e-06,
+      "loss": 0.5181,
+      "step": 20150
+    },
+    {
+      "epoch": 60.54,
+      "grad_norm": 27.785924911499023,
+      "learning_rate": 7.981981981981983e-06,
+      "loss": 0.5528,
+      "step": 20160
+    },
+    {
+      "epoch": 60.57,
+      "grad_norm": 14.472647666931152,
+      "learning_rate": 7.980980980980981e-06,
+      "loss": 0.54,
+      "step": 20170
+    },
+    {
+      "epoch": 60.6,
+      "grad_norm": 12.095008850097656,
+      "learning_rate": 7.97997997997998e-06,
+      "loss": 0.5601,
+      "step": 20180
+    },
+    {
+      "epoch": 60.63,
+      "grad_norm": 11.346393585205078,
+      "learning_rate": 7.97897897897898e-06,
+      "loss": 0.57,
+      "step": 20190
+    },
+    {
+      "epoch": 60.66,
+      "grad_norm": 16.481164932250977,
+      "learning_rate": 7.977977977977979e-06,
+      "loss": 0.5786,
+      "step": 20200
+    },
+    {
+      "epoch": 60.69,
+      "grad_norm": 12.198641777038574,
+      "learning_rate": 7.976976976976977e-06,
+      "loss": 0.5661,
+      "step": 20210
+    },
+    {
+      "epoch": 60.72,
+      "grad_norm": 15.9042329788208,
+      "learning_rate": 7.975975975975978e-06,
+      "loss": 0.6127,
+      "step": 20220
+    },
+    {
+      "epoch": 60.75,
+      "grad_norm": 16.836502075195312,
+      "learning_rate": 7.974974974974976e-06,
+      "loss": 0.5705,
+      "step": 20230
+    },
+    {
+      "epoch": 60.78,
+      "grad_norm": 16.284013748168945,
+      "learning_rate": 7.973973973973973e-06,
+      "loss": 0.6247,
+      "step": 20240
+    },
+    {
+      "epoch": 60.81,
+      "grad_norm": 13.169670104980469,
+      "learning_rate": 7.972972972972974e-06,
+      "loss": 0.5771,
+      "step": 20250
+    },
+    {
+      "epoch": 60.84,
+      "grad_norm": 15.882311820983887,
+      "learning_rate": 7.971971971971972e-06,
+      "loss": 0.6058,
+      "step": 20260
+    },
+    {
+      "epoch": 60.87,
+      "grad_norm": 18.854999542236328,
+      "learning_rate": 7.97097097097097e-06,
+      "loss": 0.5582,
+      "step": 20270
+    },
+    {
+      "epoch": 60.9,
+      "grad_norm": 11.473094940185547,
+      "learning_rate": 7.969969969969971e-06,
+      "loss": 0.6164,
+      "step": 20280
+    },
+    {
+      "epoch": 60.93,
+      "grad_norm": 17.09153175354004,
+      "learning_rate": 7.96896896896897e-06,
+      "loss": 0.5493,
+      "step": 20290
+    },
+    {
+      "epoch": 60.96,
+      "grad_norm": 16.638715744018555,
+      "learning_rate": 7.967967967967968e-06,
+      "loss": 0.5766,
+      "step": 20300
+    },
+    {
+      "epoch": 60.99,
+      "grad_norm": 17.09829330444336,
+      "learning_rate": 7.966966966966969e-06,
+      "loss": 0.5589,
+      "step": 20310
+    },
+    {
+      "epoch": 61.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.4618603587150574,
+      "eval_runtime": 12.9435,
+      "eval_samples_per_second": 772.588,
+      "eval_steps_per_second": 3.09,
+      "step": 20313
+    },
+    {
+      "epoch": 61.02,
+      "grad_norm": 13.013632774353027,
+      "learning_rate": 7.965965965965967e-06,
+      "loss": 0.515,
+      "step": 20320
+    },
+    {
+      "epoch": 61.05,
+      "grad_norm": 19.766704559326172,
+      "learning_rate": 7.964964964964966e-06,
+      "loss": 0.551,
+      "step": 20330
+    },
+    {
+      "epoch": 61.08,
+      "grad_norm": 14.834807395935059,
+      "learning_rate": 7.963963963963964e-06,
+      "loss": 0.598,
+      "step": 20340
+    },
+    {
+      "epoch": 61.11,
+      "grad_norm": 13.760293006896973,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 0.5825,
+      "step": 20350
+    },
+    {
+      "epoch": 61.14,
+      "grad_norm": 17.49335479736328,
+      "learning_rate": 7.961961961961963e-06,
+      "loss": 0.5328,
+      "step": 20360
+    },
+    {
+      "epoch": 61.17,
+      "grad_norm": 16.027816772460938,
+      "learning_rate": 7.960960960960962e-06,
+      "loss": 0.6012,
+      "step": 20370
+    },
+    {
+      "epoch": 61.2,
+      "grad_norm": 23.449174880981445,
+      "learning_rate": 7.95995995995996e-06,
+      "loss": 0.5785,
+      "step": 20380
+    },
+    {
+      "epoch": 61.23,
+      "grad_norm": 25.767745971679688,
+      "learning_rate": 7.95895895895896e-06,
+      "loss": 0.613,
+      "step": 20390
+    },
+    {
+      "epoch": 61.26,
+      "grad_norm": 11.872919082641602,
+      "learning_rate": 7.95795795795796e-06,
+      "loss": 0.5442,
+      "step": 20400
+    },
+    {
+      "epoch": 61.29,
+      "grad_norm": 14.535460472106934,
+      "learning_rate": 7.956956956956958e-06,
+      "loss": 0.599,
+      "step": 20410
+    },
+    {
+      "epoch": 61.32,
+      "grad_norm": 14.450933456420898,
+      "learning_rate": 7.955955955955956e-06,
+      "loss": 0.5674,
+      "step": 20420
+    },
+    {
+      "epoch": 61.35,
+      "grad_norm": 16.769847869873047,
+      "learning_rate": 7.954954954954955e-06,
+      "loss": 0.5948,
+      "step": 20430
+    },
+    {
+      "epoch": 61.38,
+      "grad_norm": 21.674453735351562,
+      "learning_rate": 7.953953953953955e-06,
+      "loss": 0.5648,
+      "step": 20440
+    },
+    {
+      "epoch": 61.41,
+      "grad_norm": 17.37084197998047,
+      "learning_rate": 7.952952952952954e-06,
+      "loss": 0.6078,
+      "step": 20450
+    },
+    {
+      "epoch": 61.44,
+      "grad_norm": 14.884594917297363,
+      "learning_rate": 7.951951951951953e-06,
+      "loss": 0.5549,
+      "step": 20460
+    },
+    {
+      "epoch": 61.47,
+      "grad_norm": 22.308752059936523,
+      "learning_rate": 7.950950950950951e-06,
+      "loss": 0.5316,
+      "step": 20470
+    },
+    {
+      "epoch": 61.5,
+      "grad_norm": 15.277771949768066,
+      "learning_rate": 7.949949949949951e-06,
+      "loss": 0.6423,
+      "step": 20480
+    },
+    {
+      "epoch": 61.53,
+      "grad_norm": 12.716387748718262,
+      "learning_rate": 7.948948948948948e-06,
+      "loss": 0.6142,
+      "step": 20490
+    },
+    {
+      "epoch": 61.56,
+      "grad_norm": 21.089902877807617,
+      "learning_rate": 7.947947947947949e-06,
+      "loss": 0.564,
+      "step": 20500
+    },
+    {
+      "epoch": 61.59,
+      "grad_norm": 19.012067794799805,
+      "learning_rate": 7.946946946946947e-06,
+      "loss": 0.5945,
+      "step": 20510
+    },
+    {
+      "epoch": 61.62,
+      "grad_norm": 17.51711082458496,
+      "learning_rate": 7.945945945945946e-06,
+      "loss": 0.5803,
+      "step": 20520
+    },
+    {
+      "epoch": 61.65,
+      "grad_norm": 14.719621658325195,
+      "learning_rate": 7.944944944944946e-06,
+      "loss": 0.5965,
+      "step": 20530
+    },
+    {
+      "epoch": 61.68,
+      "grad_norm": 18.21731948852539,
+      "learning_rate": 7.943943943943945e-06,
+      "loss": 0.5845,
+      "step": 20540
+    },
+    {
+      "epoch": 61.71,
+      "grad_norm": 15.174802780151367,
+      "learning_rate": 7.942942942942943e-06,
+      "loss": 0.5575,
+      "step": 20550
+    },
+    {
+      "epoch": 61.74,
+      "grad_norm": 13.877257347106934,
+      "learning_rate": 7.941941941941944e-06,
+      "loss": 0.5519,
+      "step": 20560
+    },
+    {
+      "epoch": 61.77,
+      "grad_norm": 17.353370666503906,
+      "learning_rate": 7.940940940940942e-06,
+      "loss": 0.563,
+      "step": 20570
+    },
+    {
+      "epoch": 61.8,
+      "grad_norm": 10.737649917602539,
+      "learning_rate": 7.93993993993994e-06,
+      "loss": 0.5891,
+      "step": 20580
+    },
+    {
+      "epoch": 61.83,
+      "grad_norm": 15.70091724395752,
+      "learning_rate": 7.93893893893894e-06,
+      "loss": 0.5924,
+      "step": 20590
+    },
+    {
+      "epoch": 61.86,
+      "grad_norm": 18.718290328979492,
+      "learning_rate": 7.937937937937938e-06,
+      "loss": 0.6194,
+      "step": 20600
+    },
+    {
+      "epoch": 61.89,
+      "grad_norm": 13.471919059753418,
+      "learning_rate": 7.936936936936938e-06,
+      "loss": 0.5749,
+      "step": 20610
+    },
+    {
+      "epoch": 61.92,
+      "grad_norm": 17.65549659729004,
+      "learning_rate": 7.935935935935937e-06,
+      "loss": 0.5113,
+      "step": 20620
+    },
+    {
+      "epoch": 61.95,
+      "grad_norm": 14.194541931152344,
+      "learning_rate": 7.934934934934935e-06,
+      "loss": 0.5658,
+      "step": 20630
+    },
+    {
+      "epoch": 61.98,
+      "grad_norm": 13.701493263244629,
+      "learning_rate": 7.933933933933936e-06,
+      "loss": 0.5777,
+      "step": 20640
+    },
+    {
+      "epoch": 62.0,
+      "eval_accuracy": 0.8626,
+      "eval_loss": 0.4739953577518463,
+      "eval_runtime": 12.593,
+      "eval_samples_per_second": 794.095,
+      "eval_steps_per_second": 3.176,
+      "step": 20646
+    },
+    {
+      "epoch": 62.01,
+      "grad_norm": 16.66990852355957,
+      "learning_rate": 7.932932932932934e-06,
+      "loss": 0.5321,
+      "step": 20650
+    },
+    {
+      "epoch": 62.04,
+      "grad_norm": 16.17997169494629,
+      "learning_rate": 7.931931931931933e-06,
+      "loss": 0.553,
+      "step": 20660
+    },
+    {
+      "epoch": 62.07,
+      "grad_norm": 15.974793434143066,
+      "learning_rate": 7.930930930930931e-06,
+      "loss": 0.6005,
+      "step": 20670
+    },
+    {
+      "epoch": 62.1,
+      "grad_norm": 22.15069580078125,
+      "learning_rate": 7.92992992992993e-06,
+      "loss": 0.5662,
+      "step": 20680
+    },
+    {
+      "epoch": 62.13,
+      "grad_norm": 15.846395492553711,
+      "learning_rate": 7.928928928928929e-06,
+      "loss": 0.6265,
+      "step": 20690
+    },
+    {
+      "epoch": 62.16,
+      "grad_norm": 15.250953674316406,
+      "learning_rate": 7.927927927927929e-06,
+      "loss": 0.5739,
+      "step": 20700
+    },
+    {
+      "epoch": 62.19,
+      "grad_norm": 14.215643882751465,
+      "learning_rate": 7.926926926926928e-06,
+      "loss": 0.5919,
+      "step": 20710
+    },
+    {
+      "epoch": 62.22,
+      "grad_norm": 26.54216194152832,
+      "learning_rate": 7.925925925925926e-06,
+      "loss": 0.5703,
+      "step": 20720
+    },
+    {
+      "epoch": 62.25,
+      "grad_norm": 23.78998374938965,
+      "learning_rate": 7.924924924924926e-06,
+      "loss": 0.5604,
+      "step": 20730
+    },
+    {
+      "epoch": 62.28,
+      "grad_norm": 20.236692428588867,
+      "learning_rate": 7.923923923923923e-06,
+      "loss": 0.5878,
+      "step": 20740
+    },
+    {
+      "epoch": 62.31,
+      "grad_norm": 23.307592391967773,
+      "learning_rate": 7.922922922922924e-06,
+      "loss": 0.5362,
+      "step": 20750
+    },
+    {
+      "epoch": 62.34,
+      "grad_norm": 21.87845802307129,
+      "learning_rate": 7.921921921921922e-06,
+      "loss": 0.5588,
+      "step": 20760
+    },
+    {
+      "epoch": 62.37,
+      "grad_norm": 19.75901222229004,
+      "learning_rate": 7.92092092092092e-06,
+      "loss": 0.5118,
+      "step": 20770
+    },
+    {
+      "epoch": 62.4,
+      "grad_norm": 15.26552963256836,
+      "learning_rate": 7.919919919919921e-06,
+      "loss": 0.5601,
+      "step": 20780
+    },
+    {
+      "epoch": 62.43,
+      "grad_norm": 10.958950996398926,
+      "learning_rate": 7.91891891891892e-06,
+      "loss": 0.5518,
+      "step": 20790
+    },
+    {
+      "epoch": 62.46,
+      "grad_norm": 14.27664852142334,
+      "learning_rate": 7.917917917917918e-06,
+      "loss": 0.5905,
+      "step": 20800
+    },
+    {
+      "epoch": 62.49,
+      "grad_norm": 17.700889587402344,
+      "learning_rate": 7.916916916916919e-06,
+      "loss": 0.5595,
+      "step": 20810
+    },
+    {
+      "epoch": 62.52,
+      "grad_norm": 16.927452087402344,
+      "learning_rate": 7.915915915915915e-06,
+      "loss": 0.5627,
+      "step": 20820
+    },
+    {
+      "epoch": 62.55,
+      "grad_norm": 15.279191017150879,
+      "learning_rate": 7.914914914914916e-06,
+      "loss": 0.5965,
+      "step": 20830
+    },
+    {
+      "epoch": 62.58,
+      "grad_norm": 14.534746170043945,
+      "learning_rate": 7.913913913913914e-06,
+      "loss": 0.6307,
+      "step": 20840
+    },
+    {
+      "epoch": 62.61,
+      "grad_norm": 20.03127670288086,
+      "learning_rate": 7.912912912912913e-06,
+      "loss": 0.5817,
+      "step": 20850
+    },
+    {
+      "epoch": 62.64,
+      "grad_norm": 14.393763542175293,
+      "learning_rate": 7.911911911911913e-06,
+      "loss": 0.5687,
+      "step": 20860
+    },
+    {
+      "epoch": 62.67,
+      "grad_norm": 19.6622314453125,
+      "learning_rate": 7.910910910910912e-06,
+      "loss": 0.5802,
+      "step": 20870
+    },
+    {
+      "epoch": 62.7,
+      "grad_norm": 12.865262031555176,
+      "learning_rate": 7.90990990990991e-06,
+      "loss": 0.5928,
+      "step": 20880
+    },
+    {
+      "epoch": 62.73,
+      "grad_norm": 15.107696533203125,
+      "learning_rate": 7.90890890890891e-06,
+      "loss": 0.5349,
+      "step": 20890
+    },
+    {
+      "epoch": 62.76,
+      "grad_norm": 18.19290542602539,
+      "learning_rate": 7.90790790790791e-06,
+      "loss": 0.578,
+      "step": 20900
+    },
+    {
+      "epoch": 62.79,
+      "grad_norm": 12.389555931091309,
+      "learning_rate": 7.906906906906908e-06,
+      "loss": 0.6213,
+      "step": 20910
+    },
+    {
+      "epoch": 62.82,
+      "grad_norm": 16.094959259033203,
+      "learning_rate": 7.905905905905907e-06,
+      "loss": 0.5919,
+      "step": 20920
+    },
+    {
+      "epoch": 62.85,
+      "grad_norm": 20.8697509765625,
+      "learning_rate": 7.904904904904905e-06,
+      "loss": 0.6286,
+      "step": 20930
+    },
+    {
+      "epoch": 62.88,
+      "grad_norm": 12.95927906036377,
+      "learning_rate": 7.903903903903904e-06,
+      "loss": 0.5543,
+      "step": 20940
+    },
+    {
+      "epoch": 62.91,
+      "grad_norm": 12.200501441955566,
+      "learning_rate": 7.902902902902904e-06,
+      "loss": 0.5954,
+      "step": 20950
+    },
+    {
+      "epoch": 62.94,
+      "grad_norm": 20.14626693725586,
+      "learning_rate": 7.901901901901903e-06,
+      "loss": 0.5558,
+      "step": 20960
+    },
+    {
+      "epoch": 62.97,
+      "grad_norm": 15.596393585205078,
+      "learning_rate": 7.900900900900901e-06,
+      "loss": 0.5711,
+      "step": 20970
+    },
+    {
+      "epoch": 63.0,
+      "eval_accuracy": 0.8659,
+      "eval_loss": 0.4683513045310974,
+      "eval_runtime": 12.7882,
+      "eval_samples_per_second": 781.971,
+      "eval_steps_per_second": 3.128,
+      "step": 20979
+    },
+    {
+      "epoch": 63.0,
+      "grad_norm": 16.402956008911133,
+      "learning_rate": 7.899899899899901e-06,
+      "loss": 0.5426,
+      "step": 20980
+    },
+    {
+      "epoch": 63.03,
+      "grad_norm": 16.609580993652344,
+      "learning_rate": 7.898898898898898e-06,
+      "loss": 0.5684,
+      "step": 20990
+    },
+    {
+      "epoch": 63.06,
+      "grad_norm": 19.532304763793945,
+      "learning_rate": 7.897897897897899e-06,
+      "loss": 0.589,
+      "step": 21000
+    },
+    {
+      "epoch": 63.09,
+      "grad_norm": 11.068476676940918,
+      "learning_rate": 7.896896896896897e-06,
+      "loss": 0.5477,
+      "step": 21010
+    },
+    {
+      "epoch": 63.12,
+      "grad_norm": 12.566080093383789,
+      "learning_rate": 7.895895895895896e-06,
+      "loss": 0.5507,
+      "step": 21020
+    },
+    {
+      "epoch": 63.15,
+      "grad_norm": 12.536846160888672,
+      "learning_rate": 7.894894894894896e-06,
+      "loss": 0.5802,
+      "step": 21030
+    },
+    {
+      "epoch": 63.18,
+      "grad_norm": 20.08684539794922,
+      "learning_rate": 7.893893893893895e-06,
+      "loss": 0.5974,
+      "step": 21040
+    },
+    {
+      "epoch": 63.21,
+      "grad_norm": 14.982975959777832,
+      "learning_rate": 7.892892892892893e-06,
+      "loss": 0.5331,
+      "step": 21050
+    },
+    {
+      "epoch": 63.24,
+      "grad_norm": 15.32133674621582,
+      "learning_rate": 7.891891891891894e-06,
+      "loss": 0.5339,
+      "step": 21060
+    },
+    {
+      "epoch": 63.27,
+      "grad_norm": 16.593841552734375,
+      "learning_rate": 7.89089089089089e-06,
+      "loss": 0.5521,
+      "step": 21070
+    },
+    {
+      "epoch": 63.3,
+      "grad_norm": 11.591399192810059,
+      "learning_rate": 7.88988988988989e-06,
+      "loss": 0.5025,
+      "step": 21080
+    },
+    {
+      "epoch": 63.33,
+      "grad_norm": 16.387269973754883,
+      "learning_rate": 7.88888888888889e-06,
+      "loss": 0.6144,
+      "step": 21090
+    },
+    {
+      "epoch": 63.36,
+      "grad_norm": 14.72872543334961,
+      "learning_rate": 7.887887887887888e-06,
+      "loss": 0.5563,
+      "step": 21100
+    },
+    {
+      "epoch": 63.39,
+      "grad_norm": 16.292652130126953,
+      "learning_rate": 7.886886886886888e-06,
+      "loss": 0.5957,
+      "step": 21110
+    },
+    {
+      "epoch": 63.42,
+      "grad_norm": 11.861322402954102,
+      "learning_rate": 7.885885885885887e-06,
+      "loss": 0.6094,
+      "step": 21120
+    },
+    {
+      "epoch": 63.45,
+      "grad_norm": 18.908823013305664,
+      "learning_rate": 7.884884884884885e-06,
+      "loss": 0.5862,
+      "step": 21130
+    },
+    {
+      "epoch": 63.48,
+      "grad_norm": 15.238788604736328,
+      "learning_rate": 7.883883883883886e-06,
+      "loss": 0.6308,
+      "step": 21140
+    },
+    {
+      "epoch": 63.51,
+      "grad_norm": 13.101716041564941,
+      "learning_rate": 7.882882882882884e-06,
+      "loss": 0.5427,
+      "step": 21150
+    },
+    {
+      "epoch": 63.54,
+      "grad_norm": 14.60013198852539,
+      "learning_rate": 7.881881881881881e-06,
+      "loss": 0.5067,
+      "step": 21160
+    },
+    {
+      "epoch": 63.57,
+      "grad_norm": 16.040077209472656,
+      "learning_rate": 7.880880880880882e-06,
+      "loss": 0.618,
+      "step": 21170
+    },
+    {
+      "epoch": 63.6,
+      "grad_norm": 17.32454490661621,
+      "learning_rate": 7.87987987987988e-06,
+      "loss": 0.5619,
+      "step": 21180
+    },
+    {
+      "epoch": 63.63,
+      "grad_norm": 13.533453941345215,
+      "learning_rate": 7.878878878878879e-06,
+      "loss": 0.5796,
+      "step": 21190
+    },
+    {
+      "epoch": 63.66,
+      "grad_norm": 16.205324172973633,
+      "learning_rate": 7.877877877877879e-06,
+      "loss": 0.5986,
+      "step": 21200
+    },
+    {
+      "epoch": 63.69,
+      "grad_norm": 14.960679054260254,
+      "learning_rate": 7.876876876876878e-06,
+      "loss": 0.5975,
+      "step": 21210
+    },
+    {
+      "epoch": 63.72,
+      "grad_norm": 15.52830982208252,
+      "learning_rate": 7.875875875875876e-06,
+      "loss": 0.5616,
+      "step": 21220
+    },
+    {
+      "epoch": 63.75,
+      "grad_norm": 15.617673873901367,
+      "learning_rate": 7.874874874874877e-06,
+      "loss": 0.5214,
+      "step": 21230
+    },
+    {
+      "epoch": 63.78,
+      "grad_norm": 12.792104721069336,
+      "learning_rate": 7.873873873873873e-06,
+      "loss": 0.5831,
+      "step": 21240
+    },
+    {
+      "epoch": 63.81,
+      "grad_norm": 12.508368492126465,
+      "learning_rate": 7.872872872872874e-06,
+      "loss": 0.5702,
+      "step": 21250
+    },
+    {
+      "epoch": 63.84,
+      "grad_norm": 12.276655197143555,
+      "learning_rate": 7.871871871871872e-06,
+      "loss": 0.574,
+      "step": 21260
+    },
+    {
+      "epoch": 63.87,
+      "grad_norm": 11.787044525146484,
+      "learning_rate": 7.870870870870871e-06,
+      "loss": 0.5661,
+      "step": 21270
+    },
+    {
+      "epoch": 63.9,
+      "grad_norm": 15.9622163772583,
+      "learning_rate": 7.869869869869871e-06,
+      "loss": 0.5474,
+      "step": 21280
+    },
+    {
+      "epoch": 63.93,
+      "grad_norm": 22.360185623168945,
+      "learning_rate": 7.86886886886887e-06,
+      "loss": 0.5567,
+      "step": 21290
+    },
+    {
+      "epoch": 63.96,
+      "grad_norm": 11.864326477050781,
+      "learning_rate": 7.867867867867868e-06,
+      "loss": 0.5897,
+      "step": 21300
+    },
+    {
+      "epoch": 63.99,
+      "grad_norm": 23.33677101135254,
+      "learning_rate": 7.866866866866869e-06,
+      "loss": 0.5369,
+      "step": 21310
+    },
+    {
+      "epoch": 64.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.4654677212238312,
+      "eval_runtime": 12.5277,
+      "eval_samples_per_second": 798.231,
+      "eval_steps_per_second": 3.193,
+      "step": 21312
+    },
+    {
+      "epoch": 64.02,
+      "grad_norm": 14.384602546691895,
+      "learning_rate": 7.865865865865866e-06,
+      "loss": 0.5792,
+      "step": 21320
+    },
+    {
+      "epoch": 64.05,
+      "grad_norm": 24.945512771606445,
+      "learning_rate": 7.864864864864866e-06,
+      "loss": 0.5791,
+      "step": 21330
+    },
+    {
+      "epoch": 64.08,
+      "grad_norm": 14.134567260742188,
+      "learning_rate": 7.863863863863864e-06,
+      "loss": 0.6084,
+      "step": 21340
+    },
+    {
+      "epoch": 64.11,
+      "grad_norm": 15.435029983520508,
+      "learning_rate": 7.862862862862863e-06,
+      "loss": 0.5915,
+      "step": 21350
+    },
+    {
+      "epoch": 64.14,
+      "grad_norm": 17.37160873413086,
+      "learning_rate": 7.861861861861863e-06,
+      "loss": 0.5845,
+      "step": 21360
+    },
+    {
+      "epoch": 64.17,
+      "grad_norm": 12.681687355041504,
+      "learning_rate": 7.860860860860862e-06,
+      "loss": 0.5518,
+      "step": 21370
+    },
+    {
+      "epoch": 64.2,
+      "grad_norm": 15.757386207580566,
+      "learning_rate": 7.85985985985986e-06,
+      "loss": 0.5454,
+      "step": 21380
+    },
+    {
+      "epoch": 64.23,
+      "grad_norm": 16.375354766845703,
+      "learning_rate": 7.858858858858859e-06,
+      "loss": 0.62,
+      "step": 21390
+    },
+    {
+      "epoch": 64.26,
+      "grad_norm": 15.337791442871094,
+      "learning_rate": 7.85785785785786e-06,
+      "loss": 0.5498,
+      "step": 21400
+    },
+    {
+      "epoch": 64.29,
+      "grad_norm": 13.011787414550781,
+      "learning_rate": 7.856856856856856e-06,
+      "loss": 0.5191,
+      "step": 21410
+    },
+    {
+      "epoch": 64.32,
+      "grad_norm": 14.646492958068848,
+      "learning_rate": 7.855855855855857e-06,
+      "loss": 0.5602,
+      "step": 21420
+    },
+    {
+      "epoch": 64.35,
+      "grad_norm": 14.467238426208496,
+      "learning_rate": 7.854854854854855e-06,
+      "loss": 0.5091,
+      "step": 21430
+    },
+    {
+      "epoch": 64.38,
+      "grad_norm": 16.313474655151367,
+      "learning_rate": 7.853853853853854e-06,
+      "loss": 0.5858,
+      "step": 21440
+    },
+    {
+      "epoch": 64.41,
+      "grad_norm": 15.73123550415039,
+      "learning_rate": 7.852852852852854e-06,
+      "loss": 0.5917,
+      "step": 21450
+    },
+    {
+      "epoch": 64.44,
+      "grad_norm": 17.798755645751953,
+      "learning_rate": 7.851851851851853e-06,
+      "loss": 0.5632,
+      "step": 21460
+    },
+    {
+      "epoch": 64.47,
+      "grad_norm": 13.452340126037598,
+      "learning_rate": 7.850850850850851e-06,
+      "loss": 0.5486,
+      "step": 21470
+    },
+    {
+      "epoch": 64.5,
+      "grad_norm": 12.10430908203125,
+      "learning_rate": 7.849849849849852e-06,
+      "loss": 0.5004,
+      "step": 21480
+    },
+    {
+      "epoch": 64.53,
+      "grad_norm": 16.858976364135742,
+      "learning_rate": 7.848848848848848e-06,
+      "loss": 0.5309,
+      "step": 21490
+    },
+    {
+      "epoch": 64.56,
+      "grad_norm": 16.18703269958496,
+      "learning_rate": 7.847847847847849e-06,
+      "loss": 0.6153,
+      "step": 21500
+    },
+    {
+      "epoch": 64.59,
+      "grad_norm": 19.76614761352539,
+      "learning_rate": 7.846846846846847e-06,
+      "loss": 0.6109,
+      "step": 21510
+    },
+    {
+      "epoch": 64.62,
+      "grad_norm": 15.78753662109375,
+      "learning_rate": 7.845845845845846e-06,
+      "loss": 0.5379,
+      "step": 21520
+    },
+    {
+      "epoch": 64.65,
+      "grad_norm": 15.919106483459473,
+      "learning_rate": 7.844844844844846e-06,
+      "loss": 0.5909,
+      "step": 21530
+    },
+    {
+      "epoch": 64.68,
+      "grad_norm": 16.947357177734375,
+      "learning_rate": 7.843843843843845e-06,
+      "loss": 0.5273,
+      "step": 21540
+    },
+    {
+      "epoch": 64.71,
+      "grad_norm": 14.03024959564209,
+      "learning_rate": 7.842842842842843e-06,
+      "loss": 0.5747,
+      "step": 21550
+    },
+    {
+      "epoch": 64.74,
+      "grad_norm": 14.933127403259277,
+      "learning_rate": 7.841841841841844e-06,
+      "loss": 0.4973,
+      "step": 21560
+    },
+    {
+      "epoch": 64.77,
+      "grad_norm": 15.095874786376953,
+      "learning_rate": 7.84084084084084e-06,
+      "loss": 0.5504,
+      "step": 21570
+    },
+    {
+      "epoch": 64.8,
+      "grad_norm": 14.683979034423828,
+      "learning_rate": 7.839839839839841e-06,
+      "loss": 0.5498,
+      "step": 21580
+    },
+    {
+      "epoch": 64.83,
+      "grad_norm": 17.910722732543945,
+      "learning_rate": 7.83883883883884e-06,
+      "loss": 0.5901,
+      "step": 21590
+    },
+    {
+      "epoch": 64.86,
+      "grad_norm": 13.064376831054688,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.6135,
+      "step": 21600
+    },
+    {
+      "epoch": 64.89,
+      "grad_norm": 14.728261947631836,
+      "learning_rate": 7.836836836836837e-06,
+      "loss": 0.6248,
+      "step": 21610
+    },
+    {
+      "epoch": 64.92,
+      "grad_norm": 17.558917999267578,
+      "learning_rate": 7.835835835835837e-06,
+      "loss": 0.589,
+      "step": 21620
+    },
+    {
+      "epoch": 64.95,
+      "grad_norm": 16.495952606201172,
+      "learning_rate": 7.834834834834836e-06,
+      "loss": 0.543,
+      "step": 21630
+    },
+    {
+      "epoch": 64.98,
+      "grad_norm": 24.232667922973633,
+      "learning_rate": 7.833833833833834e-06,
+      "loss": 0.5454,
+      "step": 21640
+    },
+    {
+      "epoch": 65.0,
+      "eval_accuracy": 0.867,
+      "eval_loss": 0.45735976099967957,
+      "eval_runtime": 12.4469,
+      "eval_samples_per_second": 803.414,
+      "eval_steps_per_second": 3.214,
+      "step": 21645
+    },
+    {
+      "epoch": 65.02,
+      "grad_norm": 20.44569969177246,
+      "learning_rate": 7.832832832832834e-06,
+      "loss": 0.5781,
+      "step": 21650
+    },
+    {
+      "epoch": 65.05,
+      "grad_norm": 14.24367618560791,
+      "learning_rate": 7.831831831831831e-06,
+      "loss": 0.5046,
+      "step": 21660
+    },
+    {
+      "epoch": 65.08,
+      "grad_norm": 18.760143280029297,
+      "learning_rate": 7.830830830830832e-06,
+      "loss": 0.5318,
+      "step": 21670
+    },
+    {
+      "epoch": 65.11,
+      "grad_norm": 9.44014835357666,
+      "learning_rate": 7.82982982982983e-06,
+      "loss": 0.545,
+      "step": 21680
+    },
+    {
+      "epoch": 65.14,
+      "grad_norm": 13.003317832946777,
+      "learning_rate": 7.828828828828829e-06,
+      "loss": 0.5661,
+      "step": 21690
+    },
+    {
+      "epoch": 65.17,
+      "grad_norm": 12.037182807922363,
+      "learning_rate": 7.827827827827829e-06,
+      "loss": 0.5858,
+      "step": 21700
+    },
+    {
+      "epoch": 65.2,
+      "grad_norm": 16.40542984008789,
+      "learning_rate": 7.826826826826828e-06,
+      "loss": 0.5875,
+      "step": 21710
+    },
+    {
+      "epoch": 65.23,
+      "grad_norm": 23.315446853637695,
+      "learning_rate": 7.825825825825826e-06,
+      "loss": 0.5395,
+      "step": 21720
+    },
+    {
+      "epoch": 65.26,
+      "grad_norm": 15.518813133239746,
+      "learning_rate": 7.824824824824827e-06,
+      "loss": 0.5867,
+      "step": 21730
+    },
+    {
+      "epoch": 65.29,
+      "grad_norm": 12.166085243225098,
+      "learning_rate": 7.823823823823823e-06,
+      "loss": 0.5756,
+      "step": 21740
+    },
+    {
+      "epoch": 65.32,
+      "grad_norm": 18.248538970947266,
+      "learning_rate": 7.822822822822824e-06,
+      "loss": 0.5688,
+      "step": 21750
+    },
+    {
+      "epoch": 65.35,
+      "grad_norm": 18.594797134399414,
+      "learning_rate": 7.821821821821822e-06,
+      "loss": 0.5927,
+      "step": 21760
+    },
+    {
+      "epoch": 65.38,
+      "grad_norm": 12.831320762634277,
+      "learning_rate": 7.820820820820821e-06,
+      "loss": 0.5372,
+      "step": 21770
+    },
+    {
+      "epoch": 65.41,
+      "grad_norm": 13.435611724853516,
+      "learning_rate": 7.819819819819821e-06,
+      "loss": 0.5832,
+      "step": 21780
+    },
+    {
+      "epoch": 65.44,
+      "grad_norm": 13.874771118164062,
+      "learning_rate": 7.81881881881882e-06,
+      "loss": 0.5114,
+      "step": 21790
+    },
+    {
+      "epoch": 65.47,
+      "grad_norm": 14.710037231445312,
+      "learning_rate": 7.817817817817818e-06,
+      "loss": 0.5079,
+      "step": 21800
+    },
+    {
+      "epoch": 65.5,
+      "grad_norm": 17.372941970825195,
+      "learning_rate": 7.816816816816819e-06,
+      "loss": 0.5225,
+      "step": 21810
+    },
+    {
+      "epoch": 65.53,
+      "grad_norm": 16.734004974365234,
+      "learning_rate": 7.815815815815816e-06,
+      "loss": 0.524,
+      "step": 21820
+    },
+    {
+      "epoch": 65.56,
+      "grad_norm": 17.2759952545166,
+      "learning_rate": 7.814814814814816e-06,
+      "loss": 0.5102,
+      "step": 21830
+    },
+    {
+      "epoch": 65.59,
+      "grad_norm": 12.53348445892334,
+      "learning_rate": 7.813813813813815e-06,
+      "loss": 0.5968,
+      "step": 21840
+    },
+    {
+      "epoch": 65.62,
+      "grad_norm": 15.117351531982422,
+      "learning_rate": 7.812812812812813e-06,
+      "loss": 0.5649,
+      "step": 21850
+    },
+    {
+      "epoch": 65.65,
+      "grad_norm": 14.020438194274902,
+      "learning_rate": 7.811811811811812e-06,
+      "loss": 0.5632,
+      "step": 21860
+    },
+    {
+      "epoch": 65.68,
+      "grad_norm": 19.232515335083008,
+      "learning_rate": 7.810810810810812e-06,
+      "loss": 0.5373,
+      "step": 21870
+    },
+    {
+      "epoch": 65.71,
+      "grad_norm": 18.334613800048828,
+      "learning_rate": 7.80980980980981e-06,
+      "loss": 0.5736,
+      "step": 21880
+    },
+    {
+      "epoch": 65.74,
+      "grad_norm": 14.187418937683105,
+      "learning_rate": 7.80880880880881e-06,
+      "loss": 0.5321,
+      "step": 21890
+    },
+    {
+      "epoch": 65.77,
+      "grad_norm": 23.7889461517334,
+      "learning_rate": 7.807807807807808e-06,
+      "loss": 0.5544,
+      "step": 21900
+    },
+    {
+      "epoch": 65.8,
+      "grad_norm": 13.015482902526855,
+      "learning_rate": 7.806806806806806e-06,
+      "loss": 0.6049,
+      "step": 21910
+    },
+    {
+      "epoch": 65.83,
+      "grad_norm": 19.774629592895508,
+      "learning_rate": 7.805805805805807e-06,
+      "loss": 0.5379,
+      "step": 21920
+    },
+    {
+      "epoch": 65.86,
+      "grad_norm": 17.81984519958496,
+      "learning_rate": 7.804804804804805e-06,
+      "loss": 0.6127,
+      "step": 21930
+    },
+    {
+      "epoch": 65.89,
+      "grad_norm": 11.629871368408203,
+      "learning_rate": 7.803803803803804e-06,
+      "loss": 0.5709,
+      "step": 21940
+    },
+    {
+      "epoch": 65.92,
+      "grad_norm": 14.506510734558105,
+      "learning_rate": 7.802802802802804e-06,
+      "loss": 0.5617,
+      "step": 21950
+    },
+    {
+      "epoch": 65.95,
+      "grad_norm": 13.262490272521973,
+      "learning_rate": 7.801801801801803e-06,
+      "loss": 0.6193,
+      "step": 21960
+    },
+    {
+      "epoch": 65.98,
+      "grad_norm": 16.1383113861084,
+      "learning_rate": 7.800800800800801e-06,
+      "loss": 0.5471,
+      "step": 21970
+    },
+    {
+      "epoch": 66.0,
+      "eval_accuracy": 0.8655,
+      "eval_loss": 0.4578593671321869,
+      "eval_runtime": 12.6482,
+      "eval_samples_per_second": 790.628,
+      "eval_steps_per_second": 3.163,
+      "step": 21978
+    },
+    {
+      "epoch": 66.01,
+      "grad_norm": 14.458390235900879,
+      "learning_rate": 7.799799799799802e-06,
+      "loss": 0.6371,
+      "step": 21980
+    },
+    {
+      "epoch": 66.04,
+      "grad_norm": 13.121495246887207,
+      "learning_rate": 7.798798798798799e-06,
+      "loss": 0.5389,
+      "step": 21990
+    },
+    {
+      "epoch": 66.07,
+      "grad_norm": 14.864715576171875,
+      "learning_rate": 7.797797797797799e-06,
+      "loss": 0.5968,
+      "step": 22000
+    },
+    {
+      "epoch": 66.1,
+      "grad_norm": 10.540728569030762,
+      "learning_rate": 7.796796796796797e-06,
+      "loss": 0.5739,
+      "step": 22010
+    },
+    {
+      "epoch": 66.13,
+      "grad_norm": 20.390697479248047,
+      "learning_rate": 7.795795795795796e-06,
+      "loss": 0.5319,
+      "step": 22020
+    },
+    {
+      "epoch": 66.16,
+      "grad_norm": 19.64850616455078,
+      "learning_rate": 7.794794794794796e-06,
+      "loss": 0.5881,
+      "step": 22030
+    },
+    {
+      "epoch": 66.19,
+      "grad_norm": 22.785123825073242,
+      "learning_rate": 7.793793793793795e-06,
+      "loss": 0.5249,
+      "step": 22040
+    },
+    {
+      "epoch": 66.22,
+      "grad_norm": 17.400972366333008,
+      "learning_rate": 7.792792792792793e-06,
+      "loss": 0.5222,
+      "step": 22050
+    },
+    {
+      "epoch": 66.25,
+      "grad_norm": 17.724403381347656,
+      "learning_rate": 7.791791791791792e-06,
+      "loss": 0.6046,
+      "step": 22060
+    },
+    {
+      "epoch": 66.28,
+      "grad_norm": 10.515296936035156,
+      "learning_rate": 7.79079079079079e-06,
+      "loss": 0.5991,
+      "step": 22070
+    },
+    {
+      "epoch": 66.31,
+      "grad_norm": 19.60765266418457,
+      "learning_rate": 7.78978978978979e-06,
+      "loss": 0.5919,
+      "step": 22080
+    },
+    {
+      "epoch": 66.34,
+      "grad_norm": 20.284080505371094,
+      "learning_rate": 7.78878878878879e-06,
+      "loss": 0.5722,
+      "step": 22090
+    },
+    {
+      "epoch": 66.37,
+      "grad_norm": 18.877843856811523,
+      "learning_rate": 7.787787787787788e-06,
+      "loss": 0.5648,
+      "step": 22100
+    },
+    {
+      "epoch": 66.4,
+      "grad_norm": 15.429301261901855,
+      "learning_rate": 7.786786786786787e-06,
+      "loss": 0.5848,
+      "step": 22110
+    },
+    {
+      "epoch": 66.43,
+      "grad_norm": 17.803064346313477,
+      "learning_rate": 7.785785785785787e-06,
+      "loss": 0.5631,
+      "step": 22120
+    },
+    {
+      "epoch": 66.46,
+      "grad_norm": 14.009722709655762,
+      "learning_rate": 7.784784784784786e-06,
+      "loss": 0.563,
+      "step": 22130
+    },
+    {
+      "epoch": 66.49,
+      "grad_norm": 13.050360679626465,
+      "learning_rate": 7.783783783783784e-06,
+      "loss": 0.5608,
+      "step": 22140
+    },
+    {
+      "epoch": 66.52,
+      "grad_norm": 14.04554557800293,
+      "learning_rate": 7.782782782782783e-06,
+      "loss": 0.5285,
+      "step": 22150
+    },
+    {
+      "epoch": 66.55,
+      "grad_norm": 14.689785957336426,
+      "learning_rate": 7.781781781781781e-06,
+      "loss": 0.5928,
+      "step": 22160
+    },
+    {
+      "epoch": 66.58,
+      "grad_norm": 18.265892028808594,
+      "learning_rate": 7.780780780780782e-06,
+      "loss": 0.5736,
+      "step": 22170
+    },
+    {
+      "epoch": 66.61,
+      "grad_norm": 12.933117866516113,
+      "learning_rate": 7.77977977977978e-06,
+      "loss": 0.5685,
+      "step": 22180
+    },
+    {
+      "epoch": 66.64,
+      "grad_norm": 17.895145416259766,
+      "learning_rate": 7.778778778778779e-06,
+      "loss": 0.5729,
+      "step": 22190
+    },
+    {
+      "epoch": 66.67,
+      "grad_norm": 19.048784255981445,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.6241,
+      "step": 22200
+    },
+    {
+      "epoch": 66.7,
+      "grad_norm": 12.441827774047852,
+      "learning_rate": 7.776776776776778e-06,
+      "loss": 0.5249,
+      "step": 22210
+    },
+    {
+      "epoch": 66.73,
+      "grad_norm": 12.524911880493164,
+      "learning_rate": 7.775775775775776e-06,
+      "loss": 0.5657,
+      "step": 22220
+    },
+    {
+      "epoch": 66.76,
+      "grad_norm": 25.477313995361328,
+      "learning_rate": 7.774774774774777e-06,
+      "loss": 0.5484,
+      "step": 22230
+    },
+    {
+      "epoch": 66.79,
+      "grad_norm": 15.488204002380371,
+      "learning_rate": 7.773773773773774e-06,
+      "loss": 0.5484,
+      "step": 22240
+    },
+    {
+      "epoch": 66.82,
+      "grad_norm": 12.64756965637207,
+      "learning_rate": 7.772772772772774e-06,
+      "loss": 0.4966,
+      "step": 22250
+    },
+    {
+      "epoch": 66.85,
+      "grad_norm": 20.53177261352539,
+      "learning_rate": 7.771771771771772e-06,
+      "loss": 0.6213,
+      "step": 22260
+    },
+    {
+      "epoch": 66.88,
+      "grad_norm": 12.159225463867188,
+      "learning_rate": 7.770770770770771e-06,
+      "loss": 0.6016,
+      "step": 22270
+    },
+    {
+      "epoch": 66.91,
+      "grad_norm": 20.43048858642578,
+      "learning_rate": 7.769769769769771e-06,
+      "loss": 0.5002,
+      "step": 22280
+    },
+    {
+      "epoch": 66.94,
+      "grad_norm": 16.07718276977539,
+      "learning_rate": 7.76876876876877e-06,
+      "loss": 0.612,
+      "step": 22290
+    },
+    {
+      "epoch": 66.97,
+      "grad_norm": 14.752755165100098,
+      "learning_rate": 7.767767767767769e-06,
+      "loss": 0.5301,
+      "step": 22300
+    },
+    {
+      "epoch": 67.0,
+      "grad_norm": 13.626973152160645,
+      "learning_rate": 7.766766766766767e-06,
+      "loss": 0.5816,
+      "step": 22310
+    },
+    {
+      "epoch": 67.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.46097758412361145,
+      "eval_runtime": 12.9948,
+      "eval_samples_per_second": 769.54,
+      "eval_steps_per_second": 3.078,
+      "step": 22311
+    },
+    {
+      "epoch": 67.03,
+      "grad_norm": 17.61612892150879,
+      "learning_rate": 7.765765765765766e-06,
+      "loss": 0.5121,
+      "step": 22320
+    },
+    {
+      "epoch": 67.06,
+      "grad_norm": 14.522309303283691,
+      "learning_rate": 7.764764764764764e-06,
+      "loss": 0.5258,
+      "step": 22330
+    },
+    {
+      "epoch": 67.09,
+      "grad_norm": 12.155220031738281,
+      "learning_rate": 7.763763763763765e-06,
+      "loss": 0.5355,
+      "step": 22340
+    },
+    {
+      "epoch": 67.12,
+      "grad_norm": 17.316837310791016,
+      "learning_rate": 7.762762762762763e-06,
+      "loss": 0.5247,
+      "step": 22350
+    },
+    {
+      "epoch": 67.15,
+      "grad_norm": 20.70302963256836,
+      "learning_rate": 7.761761761761762e-06,
+      "loss": 0.5918,
+      "step": 22360
+    },
+    {
+      "epoch": 67.18,
+      "grad_norm": 15.649313926696777,
+      "learning_rate": 7.760760760760762e-06,
+      "loss": 0.5231,
+      "step": 22370
+    },
+    {
+      "epoch": 67.21,
+      "grad_norm": 12.764701843261719,
+      "learning_rate": 7.75975975975976e-06,
+      "loss": 0.5763,
+      "step": 22380
+    },
+    {
+      "epoch": 67.24,
+      "grad_norm": 22.652385711669922,
+      "learning_rate": 7.75875875875876e-06,
+      "loss": 0.5285,
+      "step": 22390
+    },
+    {
+      "epoch": 67.27,
+      "grad_norm": 20.876888275146484,
+      "learning_rate": 7.757757757757758e-06,
+      "loss": 0.6043,
+      "step": 22400
+    },
+    {
+      "epoch": 67.3,
+      "grad_norm": 16.850923538208008,
+      "learning_rate": 7.756756756756756e-06,
+      "loss": 0.6162,
+      "step": 22410
+    },
+    {
+      "epoch": 67.33,
+      "grad_norm": 12.737711906433105,
+      "learning_rate": 7.755755755755757e-06,
+      "loss": 0.5823,
+      "step": 22420
+    },
+    {
+      "epoch": 67.36,
+      "grad_norm": 15.0642728805542,
+      "learning_rate": 7.754754754754755e-06,
+      "loss": 0.5225,
+      "step": 22430
+    },
+    {
+      "epoch": 67.39,
+      "grad_norm": 14.839635848999023,
+      "learning_rate": 7.753753753753754e-06,
+      "loss": 0.572,
+      "step": 22440
+    },
+    {
+      "epoch": 67.42,
+      "grad_norm": 15.499557495117188,
+      "learning_rate": 7.752752752752754e-06,
+      "loss": 0.5427,
+      "step": 22450
+    },
+    {
+      "epoch": 67.45,
+      "grad_norm": 21.27898597717285,
+      "learning_rate": 7.751751751751753e-06,
+      "loss": 0.5442,
+      "step": 22460
+    },
+    {
+      "epoch": 67.48,
+      "grad_norm": 18.203609466552734,
+      "learning_rate": 7.750750750750751e-06,
+      "loss": 0.5685,
+      "step": 22470
+    },
+    {
+      "epoch": 67.51,
+      "grad_norm": 14.603145599365234,
+      "learning_rate": 7.749749749749752e-06,
+      "loss": 0.6034,
+      "step": 22480
+    },
+    {
+      "epoch": 67.54,
+      "grad_norm": 19.590543746948242,
+      "learning_rate": 7.748748748748749e-06,
+      "loss": 0.5384,
+      "step": 22490
+    },
+    {
+      "epoch": 67.57,
+      "grad_norm": 17.51311683654785,
+      "learning_rate": 7.747747747747749e-06,
+      "loss": 0.5626,
+      "step": 22500
+    },
+    {
+      "epoch": 67.6,
+      "grad_norm": 14.876330375671387,
+      "learning_rate": 7.746746746746747e-06,
+      "loss": 0.5592,
+      "step": 22510
+    },
+    {
+      "epoch": 67.63,
+      "grad_norm": 22.465009689331055,
+      "learning_rate": 7.745745745745746e-06,
+      "loss": 0.5905,
+      "step": 22520
+    },
+    {
+      "epoch": 67.66,
+      "grad_norm": 14.902283668518066,
+      "learning_rate": 7.744744744744745e-06,
+      "loss": 0.5325,
+      "step": 22530
+    },
+    {
+      "epoch": 67.69,
+      "grad_norm": 12.203268051147461,
+      "learning_rate": 7.743743743743745e-06,
+      "loss": 0.482,
+      "step": 22540
+    },
+    {
+      "epoch": 67.72,
+      "grad_norm": 17.009416580200195,
+      "learning_rate": 7.742742742742744e-06,
+      "loss": 0.5318,
+      "step": 22550
+    },
+    {
+      "epoch": 67.75,
+      "grad_norm": 32.27000427246094,
+      "learning_rate": 7.741741741741742e-06,
+      "loss": 0.5071,
+      "step": 22560
+    },
+    {
+      "epoch": 67.78,
+      "grad_norm": 14.510608673095703,
+      "learning_rate": 7.74074074074074e-06,
+      "loss": 0.5233,
+      "step": 22570
+    },
+    {
+      "epoch": 67.81,
+      "grad_norm": 16.449337005615234,
+      "learning_rate": 7.73973973973974e-06,
+      "loss": 0.4803,
+      "step": 22580
+    },
+    {
+      "epoch": 67.84,
+      "grad_norm": 19.513935089111328,
+      "learning_rate": 7.73873873873874e-06,
+      "loss": 0.6083,
+      "step": 22590
+    },
+    {
+      "epoch": 67.87,
+      "grad_norm": 17.661775588989258,
+      "learning_rate": 7.737737737737738e-06,
+      "loss": 0.5604,
+      "step": 22600
+    },
+    {
+      "epoch": 67.9,
+      "grad_norm": 19.88722038269043,
+      "learning_rate": 7.736736736736737e-06,
+      "loss": 0.5425,
+      "step": 22610
+    },
+    {
+      "epoch": 67.93,
+      "grad_norm": 21.868816375732422,
+      "learning_rate": 7.735735735735737e-06,
+      "loss": 0.5772,
+      "step": 22620
+    },
+    {
+      "epoch": 67.96,
+      "grad_norm": 17.84311294555664,
+      "learning_rate": 7.734734734734736e-06,
+      "loss": 0.5472,
+      "step": 22630
+    },
+    {
+      "epoch": 67.99,
+      "grad_norm": 10.967425346374512,
+      "learning_rate": 7.733733733733734e-06,
+      "loss": 0.5262,
+      "step": 22640
+    },
+    {
+      "epoch": 68.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.4631381928920746,
+      "eval_runtime": 12.8581,
+      "eval_samples_per_second": 777.719,
+      "eval_steps_per_second": 3.111,
+      "step": 22644
+    },
+    {
+      "epoch": 68.02,
+      "grad_norm": 25.05714225769043,
+      "learning_rate": 7.732732732732733e-06,
+      "loss": 0.5734,
+      "step": 22650
+    },
+    {
+      "epoch": 68.05,
+      "grad_norm": 15.997255325317383,
+      "learning_rate": 7.731731731731731e-06,
+      "loss": 0.5517,
+      "step": 22660
+    },
+    {
+      "epoch": 68.08,
+      "grad_norm": 14.824918746948242,
+      "learning_rate": 7.730730730730732e-06,
+      "loss": 0.5681,
+      "step": 22670
+    },
+    {
+      "epoch": 68.11,
+      "grad_norm": 23.816343307495117,
+      "learning_rate": 7.72972972972973e-06,
+      "loss": 0.5502,
+      "step": 22680
+    },
+    {
+      "epoch": 68.14,
+      "grad_norm": 18.046607971191406,
+      "learning_rate": 7.728728728728729e-06,
+      "loss": 0.5043,
+      "step": 22690
+    },
+    {
+      "epoch": 68.17,
+      "grad_norm": 15.784799575805664,
+      "learning_rate": 7.72772772772773e-06,
+      "loss": 0.5441,
+      "step": 22700
+    },
+    {
+      "epoch": 68.2,
+      "grad_norm": 11.996554374694824,
+      "learning_rate": 7.726726726726728e-06,
+      "loss": 0.5802,
+      "step": 22710
+    },
+    {
+      "epoch": 68.23,
+      "grad_norm": 13.040654182434082,
+      "learning_rate": 7.725725725725726e-06,
+      "loss": 0.5738,
+      "step": 22720
+    },
+    {
+      "epoch": 68.26,
+      "grad_norm": 18.27982521057129,
+      "learning_rate": 7.724724724724727e-06,
+      "loss": 0.5618,
+      "step": 22730
+    },
+    {
+      "epoch": 68.29,
+      "grad_norm": 16.63296127319336,
+      "learning_rate": 7.723723723723724e-06,
+      "loss": 0.5308,
+      "step": 22740
+    },
+    {
+      "epoch": 68.32,
+      "grad_norm": 15.186198234558105,
+      "learning_rate": 7.722722722722722e-06,
+      "loss": 0.5959,
+      "step": 22750
+    },
+    {
+      "epoch": 68.35,
+      "grad_norm": 19.347002029418945,
+      "learning_rate": 7.721721721721722e-06,
+      "loss": 0.5802,
+      "step": 22760
+    },
+    {
+      "epoch": 68.38,
+      "grad_norm": 15.88683032989502,
+      "learning_rate": 7.720720720720721e-06,
+      "loss": 0.5528,
+      "step": 22770
+    },
+    {
+      "epoch": 68.41,
+      "grad_norm": 15.911577224731445,
+      "learning_rate": 7.71971971971972e-06,
+      "loss": 0.5677,
+      "step": 22780
+    },
+    {
+      "epoch": 68.44,
+      "grad_norm": 11.782918930053711,
+      "learning_rate": 7.71871871871872e-06,
+      "loss": 0.5572,
+      "step": 22790
+    },
+    {
+      "epoch": 68.47,
+      "grad_norm": 14.087546348571777,
+      "learning_rate": 7.717717717717719e-06,
+      "loss": 0.5194,
+      "step": 22800
+    },
+    {
+      "epoch": 68.5,
+      "grad_norm": 16.74912452697754,
+      "learning_rate": 7.716716716716717e-06,
+      "loss": 0.5816,
+      "step": 22810
+    },
+    {
+      "epoch": 68.53,
+      "grad_norm": 16.07416534423828,
+      "learning_rate": 7.715715715715716e-06,
+      "loss": 0.5926,
+      "step": 22820
+    },
+    {
+      "epoch": 68.56,
+      "grad_norm": 13.526765823364258,
+      "learning_rate": 7.714714714714714e-06,
+      "loss": 0.528,
+      "step": 22830
+    },
+    {
+      "epoch": 68.59,
+      "grad_norm": 15.224496841430664,
+      "learning_rate": 7.713713713713715e-06,
+      "loss": 0.5792,
+      "step": 22840
+    },
+    {
+      "epoch": 68.62,
+      "grad_norm": 13.790152549743652,
+      "learning_rate": 7.712712712712713e-06,
+      "loss": 0.6411,
+      "step": 22850
+    },
+    {
+      "epoch": 68.65,
+      "grad_norm": 14.932406425476074,
+      "learning_rate": 7.711711711711712e-06,
+      "loss": 0.5628,
+      "step": 22860
+    },
+    {
+      "epoch": 68.68,
+      "grad_norm": 15.2655668258667,
+      "learning_rate": 7.710710710710712e-06,
+      "loss": 0.5397,
+      "step": 22870
+    },
+    {
+      "epoch": 68.71,
+      "grad_norm": 19.84419822692871,
+      "learning_rate": 7.70970970970971e-06,
+      "loss": 0.5996,
+      "step": 22880
+    },
+    {
+      "epoch": 68.74,
+      "grad_norm": 15.053720474243164,
+      "learning_rate": 7.70870870870871e-06,
+      "loss": 0.532,
+      "step": 22890
+    },
+    {
+      "epoch": 68.77,
+      "grad_norm": 13.848394393920898,
+      "learning_rate": 7.707707707707708e-06,
+      "loss": 0.4833,
+      "step": 22900
+    },
+    {
+      "epoch": 68.8,
+      "grad_norm": 18.321563720703125,
+      "learning_rate": 7.706706706706707e-06,
+      "loss": 0.5969,
+      "step": 22910
+    },
+    {
+      "epoch": 68.83,
+      "grad_norm": 17.008100509643555,
+      "learning_rate": 7.705705705705707e-06,
+      "loss": 0.5224,
+      "step": 22920
+    },
+    {
+      "epoch": 68.86,
+      "grad_norm": 9.902338981628418,
+      "learning_rate": 7.704704704704705e-06,
+      "loss": 0.5066,
+      "step": 22930
+    },
+    {
+      "epoch": 68.89,
+      "grad_norm": 19.7772159576416,
+      "learning_rate": 7.703703703703704e-06,
+      "loss": 0.5297,
+      "step": 22940
+    },
+    {
+      "epoch": 68.92,
+      "grad_norm": 17.880611419677734,
+      "learning_rate": 7.702702702702704e-06,
+      "loss": 0.5247,
+      "step": 22950
+    },
+    {
+      "epoch": 68.95,
+      "grad_norm": 12.586747169494629,
+      "learning_rate": 7.701701701701703e-06,
+      "loss": 0.5011,
+      "step": 22960
+    },
+    {
+      "epoch": 68.98,
+      "grad_norm": 18.35675621032715,
+      "learning_rate": 7.700700700700701e-06,
+      "loss": 0.5163,
+      "step": 22970
+    },
+    {
+      "epoch": 69.0,
+      "eval_accuracy": 0.8677,
+      "eval_loss": 0.45321527123451233,
+      "eval_runtime": 12.8451,
+      "eval_samples_per_second": 778.506,
+      "eval_steps_per_second": 3.114,
+      "step": 22977
+    },
+    {
+      "epoch": 69.01,
+      "grad_norm": 13.263096809387207,
+      "learning_rate": 7.6996996996997e-06,
+      "loss": 0.5317,
+      "step": 22980
+    },
+    {
+      "epoch": 69.04,
+      "grad_norm": 22.147701263427734,
+      "learning_rate": 7.698698698698699e-06,
+      "loss": 0.5439,
+      "step": 22990
+    },
+    {
+      "epoch": 69.07,
+      "grad_norm": 18.904695510864258,
+      "learning_rate": 7.697697697697697e-06,
+      "loss": 0.5396,
+      "step": 23000
+    },
+    {
+      "epoch": 69.1,
+      "grad_norm": 14.03147029876709,
+      "learning_rate": 7.696696696696698e-06,
+      "loss": 0.5593,
+      "step": 23010
+    },
+    {
+      "epoch": 69.13,
+      "grad_norm": 11.370265007019043,
+      "learning_rate": 7.695695695695696e-06,
+      "loss": 0.5129,
+      "step": 23020
+    },
+    {
+      "epoch": 69.16,
+      "grad_norm": 13.232316017150879,
+      "learning_rate": 7.694694694694695e-06,
+      "loss": 0.573,
+      "step": 23030
+    },
+    {
+      "epoch": 69.19,
+      "grad_norm": 19.535900115966797,
+      "learning_rate": 7.693693693693695e-06,
+      "loss": 0.5484,
+      "step": 23040
+    },
+    {
+      "epoch": 69.22,
+      "grad_norm": 22.441991806030273,
+      "learning_rate": 7.692692692692694e-06,
+      "loss": 0.5777,
+      "step": 23050
+    },
+    {
+      "epoch": 69.25,
+      "grad_norm": 14.082656860351562,
+      "learning_rate": 7.691691691691692e-06,
+      "loss": 0.4827,
+      "step": 23060
+    },
+    {
+      "epoch": 69.28,
+      "grad_norm": 18.965608596801758,
+      "learning_rate": 7.69069069069069e-06,
+      "loss": 0.5555,
+      "step": 23070
+    },
+    {
+      "epoch": 69.31,
+      "grad_norm": 21.62281608581543,
+      "learning_rate": 7.68968968968969e-06,
+      "loss": 0.4925,
+      "step": 23080
+    },
+    {
+      "epoch": 69.34,
+      "grad_norm": 13.503475189208984,
+      "learning_rate": 7.68868868868869e-06,
+      "loss": 0.5216,
+      "step": 23090
+    },
+    {
+      "epoch": 69.37,
+      "grad_norm": 18.784984588623047,
+      "learning_rate": 7.687687687687688e-06,
+      "loss": 0.526,
+      "step": 23100
+    },
+    {
+      "epoch": 69.4,
+      "grad_norm": 18.865676879882812,
+      "learning_rate": 7.686686686686687e-06,
+      "loss": 0.553,
+      "step": 23110
+    },
+    {
+      "epoch": 69.43,
+      "grad_norm": 16.839923858642578,
+      "learning_rate": 7.685685685685687e-06,
+      "loss": 0.5339,
+      "step": 23120
+    },
+    {
+      "epoch": 69.46,
+      "grad_norm": 14.643489837646484,
+      "learning_rate": 7.684684684684686e-06,
+      "loss": 0.5262,
+      "step": 23130
+    },
+    {
+      "epoch": 69.49,
+      "grad_norm": 15.389371871948242,
+      "learning_rate": 7.683683683683684e-06,
+      "loss": 0.5784,
+      "step": 23140
+    },
+    {
+      "epoch": 69.52,
+      "grad_norm": 12.793866157531738,
+      "learning_rate": 7.682682682682683e-06,
+      "loss": 0.5171,
+      "step": 23150
+    },
+    {
+      "epoch": 69.55,
+      "grad_norm": 11.249611854553223,
+      "learning_rate": 7.681681681681682e-06,
+      "loss": 0.5073,
+      "step": 23160
+    },
+    {
+      "epoch": 69.58,
+      "grad_norm": 16.497753143310547,
+      "learning_rate": 7.680680680680682e-06,
+      "loss": 0.5207,
+      "step": 23170
+    },
+    {
+      "epoch": 69.61,
+      "grad_norm": 16.578771591186523,
+      "learning_rate": 7.67967967967968e-06,
+      "loss": 0.5069,
+      "step": 23180
+    },
+    {
+      "epoch": 69.64,
+      "grad_norm": 17.70914077758789,
+      "learning_rate": 7.678678678678679e-06,
+      "loss": 0.5507,
+      "step": 23190
+    },
+    {
+      "epoch": 69.67,
+      "grad_norm": 14.014601707458496,
+      "learning_rate": 7.67767767767768e-06,
+      "loss": 0.5078,
+      "step": 23200
+    },
+    {
+      "epoch": 69.7,
+      "grad_norm": 17.132427215576172,
+      "learning_rate": 7.676676676676678e-06,
+      "loss": 0.541,
+      "step": 23210
+    },
+    {
+      "epoch": 69.73,
+      "grad_norm": 23.833972930908203,
+      "learning_rate": 7.675675675675676e-06,
+      "loss": 0.5316,
+      "step": 23220
+    },
+    {
+      "epoch": 69.76,
+      "grad_norm": 15.896707534790039,
+      "learning_rate": 7.674674674674675e-06,
+      "loss": 0.5324,
+      "step": 23230
+    },
+    {
+      "epoch": 69.79,
+      "grad_norm": 22.634628295898438,
+      "learning_rate": 7.673673673673674e-06,
+      "loss": 0.5618,
+      "step": 23240
+    },
+    {
+      "epoch": 69.82,
+      "grad_norm": 20.63096809387207,
+      "learning_rate": 7.672672672672672e-06,
+      "loss": 0.5282,
+      "step": 23250
+    },
+    {
+      "epoch": 69.85,
+      "grad_norm": 16.865144729614258,
+      "learning_rate": 7.671671671671673e-06,
+      "loss": 0.5597,
+      "step": 23260
+    },
+    {
+      "epoch": 69.88,
+      "grad_norm": 12.029297828674316,
+      "learning_rate": 7.670670670670671e-06,
+      "loss": 0.52,
+      "step": 23270
+    },
+    {
+      "epoch": 69.91,
+      "grad_norm": 19.67845344543457,
+      "learning_rate": 7.66966966966967e-06,
+      "loss": 0.582,
+      "step": 23280
+    },
+    {
+      "epoch": 69.94,
+      "grad_norm": 10.350509643554688,
+      "learning_rate": 7.66866866866867e-06,
+      "loss": 0.5364,
+      "step": 23290
+    },
+    {
+      "epoch": 69.97,
+      "grad_norm": 16.220821380615234,
+      "learning_rate": 7.667667667667669e-06,
+      "loss": 0.5446,
+      "step": 23300
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 57.43446350097656,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 0.5231,
+      "step": 23310
+    },
+    {
+      "epoch": 70.0,
+      "eval_accuracy": 0.867,
+      "eval_loss": 0.46347084641456604,
+      "eval_runtime": 13.0652,
+      "eval_samples_per_second": 765.395,
+      "eval_steps_per_second": 3.062,
+      "step": 23310
+    },
+    {
+      "epoch": 70.03,
+      "grad_norm": 15.071653366088867,
+      "learning_rate": 7.665665665665666e-06,
+      "loss": 0.5315,
+      "step": 23320
+    },
+    {
+      "epoch": 70.06,
+      "grad_norm": 33.71089553833008,
+      "learning_rate": 7.664664664664664e-06,
+      "loss": 0.5371,
+      "step": 23330
+    },
+    {
+      "epoch": 70.09,
+      "grad_norm": 18.549217224121094,
+      "learning_rate": 7.663663663663665e-06,
+      "loss": 0.5398,
+      "step": 23340
+    },
+    {
+      "epoch": 70.12,
+      "grad_norm": 19.0047550201416,
+      "learning_rate": 7.662662662662663e-06,
+      "loss": 0.5,
+      "step": 23350
+    },
+    {
+      "epoch": 70.15,
+      "grad_norm": 12.97152042388916,
+      "learning_rate": 7.661661661661662e-06,
+      "loss": 0.5523,
+      "step": 23360
+    },
+    {
+      "epoch": 70.18,
+      "grad_norm": 16.551597595214844,
+      "learning_rate": 7.660660660660662e-06,
+      "loss": 0.5394,
+      "step": 23370
+    },
+    {
+      "epoch": 70.21,
+      "grad_norm": 17.33650779724121,
+      "learning_rate": 7.65965965965966e-06,
+      "loss": 0.5095,
+      "step": 23380
+    },
+    {
+      "epoch": 70.24,
+      "grad_norm": 20.949827194213867,
+      "learning_rate": 7.65865865865866e-06,
+      "loss": 0.5707,
+      "step": 23390
+    },
+    {
+      "epoch": 70.27,
+      "grad_norm": 17.115488052368164,
+      "learning_rate": 7.657657657657658e-06,
+      "loss": 0.5053,
+      "step": 23400
+    },
+    {
+      "epoch": 70.3,
+      "grad_norm": 18.152801513671875,
+      "learning_rate": 7.656656656656657e-06,
+      "loss": 0.5563,
+      "step": 23410
+    },
+    {
+      "epoch": 70.33,
+      "grad_norm": 24.085355758666992,
+      "learning_rate": 7.655655655655657e-06,
+      "loss": 0.5389,
+      "step": 23420
+    },
+    {
+      "epoch": 70.36,
+      "grad_norm": 18.19046401977539,
+      "learning_rate": 7.654654654654655e-06,
+      "loss": 0.5402,
+      "step": 23430
+    },
+    {
+      "epoch": 70.39,
+      "grad_norm": 19.540538787841797,
+      "learning_rate": 7.653653653653654e-06,
+      "loss": 0.5509,
+      "step": 23440
+    },
+    {
+      "epoch": 70.42,
+      "grad_norm": 18.179523468017578,
+      "learning_rate": 7.652652652652653e-06,
+      "loss": 0.5563,
+      "step": 23450
+    },
+    {
+      "epoch": 70.45,
+      "grad_norm": 11.197782516479492,
+      "learning_rate": 7.651651651651653e-06,
+      "loss": 0.4984,
+      "step": 23460
+    },
+    {
+      "epoch": 70.48,
+      "grad_norm": 18.672710418701172,
+      "learning_rate": 7.650650650650652e-06,
+      "loss": 0.553,
+      "step": 23470
+    },
+    {
+      "epoch": 70.51,
+      "grad_norm": 20.223995208740234,
+      "learning_rate": 7.64964964964965e-06,
+      "loss": 0.54,
+      "step": 23480
+    },
+    {
+      "epoch": 70.54,
+      "grad_norm": 19.691944122314453,
+      "learning_rate": 7.648648648648649e-06,
+      "loss": 0.6133,
+      "step": 23490
+    },
+    {
+      "epoch": 70.57,
+      "grad_norm": 13.218948364257812,
+      "learning_rate": 7.647647647647647e-06,
+      "loss": 0.536,
+      "step": 23500
+    },
+    {
+      "epoch": 70.6,
+      "grad_norm": 18.09910774230957,
+      "learning_rate": 7.646646646646648e-06,
+      "loss": 0.5601,
+      "step": 23510
+    },
+    {
+      "epoch": 70.63,
+      "grad_norm": 15.12651252746582,
+      "learning_rate": 7.645645645645646e-06,
+      "loss": 0.5298,
+      "step": 23520
+    },
+    {
+      "epoch": 70.66,
+      "grad_norm": 13.266009330749512,
+      "learning_rate": 7.644644644644645e-06,
+      "loss": 0.5261,
+      "step": 23530
+    },
+    {
+      "epoch": 70.69,
+      "grad_norm": 15.739737510681152,
+      "learning_rate": 7.643643643643645e-06,
+      "loss": 0.5563,
+      "step": 23540
+    },
+    {
+      "epoch": 70.72,
+      "grad_norm": 19.626802444458008,
+      "learning_rate": 7.642642642642644e-06,
+      "loss": 0.5593,
+      "step": 23550
+    },
+    {
+      "epoch": 70.75,
+      "grad_norm": 15.771201133728027,
+      "learning_rate": 7.641641641641642e-06,
+      "loss": 0.4422,
+      "step": 23560
+    },
+    {
+      "epoch": 70.78,
+      "grad_norm": 12.177371978759766,
+      "learning_rate": 7.640640640640641e-06,
+      "loss": 0.5231,
+      "step": 23570
+    },
+    {
+      "epoch": 70.81,
+      "grad_norm": 17.018787384033203,
+      "learning_rate": 7.63963963963964e-06,
+      "loss": 0.5769,
+      "step": 23580
+    },
+    {
+      "epoch": 70.84,
+      "grad_norm": 13.767796516418457,
+      "learning_rate": 7.63863863863864e-06,
+      "loss": 0.5271,
+      "step": 23590
+    },
+    {
+      "epoch": 70.87,
+      "grad_norm": 13.441535949707031,
+      "learning_rate": 7.637637637637638e-06,
+      "loss": 0.5156,
+      "step": 23600
+    },
+    {
+      "epoch": 70.9,
+      "grad_norm": 8.215331077575684,
+      "learning_rate": 7.636636636636637e-06,
+      "loss": 0.506,
+      "step": 23610
+    },
+    {
+      "epoch": 70.93,
+      "grad_norm": 16.722057342529297,
+      "learning_rate": 7.635635635635637e-06,
+      "loss": 0.611,
+      "step": 23620
+    },
+    {
+      "epoch": 70.96,
+      "grad_norm": 13.369221687316895,
+      "learning_rate": 7.634634634634636e-06,
+      "loss": 0.5239,
+      "step": 23630
+    },
+    {
+      "epoch": 70.99,
+      "grad_norm": 18.794593811035156,
+      "learning_rate": 7.633633633633634e-06,
+      "loss": 0.5672,
+      "step": 23640
+    },
+    {
+      "epoch": 71.0,
+      "eval_accuracy": 0.8668,
+      "eval_loss": 0.4625888764858246,
+      "eval_runtime": 12.8291,
+      "eval_samples_per_second": 779.477,
+      "eval_steps_per_second": 3.118,
+      "step": 23643
+    },
+    {
+      "epoch": 71.02,
+      "grad_norm": 14.65259075164795,
+      "learning_rate": 7.632632632632633e-06,
+      "loss": 0.5242,
+      "step": 23650
+    },
+    {
+      "epoch": 71.05,
+      "grad_norm": 14.861987113952637,
+      "learning_rate": 7.631631631631632e-06,
+      "loss": 0.525,
+      "step": 23660
+    },
+    {
+      "epoch": 71.08,
+      "grad_norm": 17.436002731323242,
+      "learning_rate": 7.63063063063063e-06,
+      "loss": 0.5318,
+      "step": 23670
+    },
+    {
+      "epoch": 71.11,
+      "grad_norm": 16.081445693969727,
+      "learning_rate": 7.62962962962963e-06,
+      "loss": 0.5621,
+      "step": 23680
+    },
+    {
+      "epoch": 71.14,
+      "grad_norm": 14.570060729980469,
+      "learning_rate": 7.628628628628629e-06,
+      "loss": 0.545,
+      "step": 23690
+    },
+    {
+      "epoch": 71.17,
+      "grad_norm": 22.60666847229004,
+      "learning_rate": 7.6276276276276285e-06,
+      "loss": 0.5597,
+      "step": 23700
+    },
+    {
+      "epoch": 71.2,
+      "grad_norm": 15.404147148132324,
+      "learning_rate": 7.626626626626628e-06,
+      "loss": 0.5265,
+      "step": 23710
+    },
+    {
+      "epoch": 71.23,
+      "grad_norm": 14.775761604309082,
+      "learning_rate": 7.6256256256256266e-06,
+      "loss": 0.5357,
+      "step": 23720
+    },
+    {
+      "epoch": 71.26,
+      "grad_norm": 10.40296745300293,
+      "learning_rate": 7.624624624624624e-06,
+      "loss": 0.5307,
+      "step": 23730
+    },
+    {
+      "epoch": 71.29,
+      "grad_norm": 21.786359786987305,
+      "learning_rate": 7.623623623623624e-06,
+      "loss": 0.5797,
+      "step": 23740
+    },
+    {
+      "epoch": 71.32,
+      "grad_norm": 15.039482116699219,
+      "learning_rate": 7.622622622622623e-06,
+      "loss": 0.4593,
+      "step": 23750
+    },
+    {
+      "epoch": 71.35,
+      "grad_norm": 16.6463623046875,
+      "learning_rate": 7.621621621621622e-06,
+      "loss": 0.5371,
+      "step": 23760
+    },
+    {
+      "epoch": 71.38,
+      "grad_norm": 18.40976333618164,
+      "learning_rate": 7.620620620620621e-06,
+      "loss": 0.5448,
+      "step": 23770
+    },
+    {
+      "epoch": 71.41,
+      "grad_norm": 18.807512283325195,
+      "learning_rate": 7.619619619619621e-06,
+      "loss": 0.5824,
+      "step": 23780
+    },
+    {
+      "epoch": 71.44,
+      "grad_norm": 18.5842342376709,
+      "learning_rate": 7.618618618618619e-06,
+      "loss": 0.5171,
+      "step": 23790
+    },
+    {
+      "epoch": 71.47,
+      "grad_norm": 23.46646499633789,
+      "learning_rate": 7.617617617617619e-06,
+      "loss": 0.5398,
+      "step": 23800
+    },
+    {
+      "epoch": 71.5,
+      "grad_norm": 12.604337692260742,
+      "learning_rate": 7.616616616616618e-06,
+      "loss": 0.5405,
+      "step": 23810
+    },
+    {
+      "epoch": 71.53,
+      "grad_norm": 19.489194869995117,
+      "learning_rate": 7.615615615615616e-06,
+      "loss": 0.5279,
+      "step": 23820
+    },
+    {
+      "epoch": 71.56,
+      "grad_norm": 17.195911407470703,
+      "learning_rate": 7.614614614614615e-06,
+      "loss": 0.5924,
+      "step": 23830
+    },
+    {
+      "epoch": 71.59,
+      "grad_norm": 16.526309967041016,
+      "learning_rate": 7.613613613613614e-06,
+      "loss": 0.5628,
+      "step": 23840
+    },
+    {
+      "epoch": 71.62,
+      "grad_norm": 19.324848175048828,
+      "learning_rate": 7.612612612612613e-06,
+      "loss": 0.5232,
+      "step": 23850
+    },
+    {
+      "epoch": 71.65,
+      "grad_norm": 17.28827476501465,
+      "learning_rate": 7.611611611611612e-06,
+      "loss": 0.5247,
+      "step": 23860
+    },
+    {
+      "epoch": 71.68,
+      "grad_norm": 16.0518798828125,
+      "learning_rate": 7.610610610610611e-06,
+      "loss": 0.5137,
+      "step": 23870
+    },
+    {
+      "epoch": 71.71,
+      "grad_norm": 15.90246295928955,
+      "learning_rate": 7.609609609609611e-06,
+      "loss": 0.5948,
+      "step": 23880
+    },
+    {
+      "epoch": 71.74,
+      "grad_norm": 13.580601692199707,
+      "learning_rate": 7.6086086086086095e-06,
+      "loss": 0.5067,
+      "step": 23890
+    },
+    {
+      "epoch": 71.77,
+      "grad_norm": 16.561983108520508,
+      "learning_rate": 7.607607607607608e-06,
+      "loss": 0.5639,
+      "step": 23900
+    },
+    {
+      "epoch": 71.8,
+      "grad_norm": 22.43711280822754,
+      "learning_rate": 7.606606606606607e-06,
+      "loss": 0.5548,
+      "step": 23910
+    },
+    {
+      "epoch": 71.83,
+      "grad_norm": 15.016554832458496,
+      "learning_rate": 7.605605605605606e-06,
+      "loss": 0.5917,
+      "step": 23920
+    },
+    {
+      "epoch": 71.86,
+      "grad_norm": 15.802587509155273,
+      "learning_rate": 7.6046046046046055e-06,
+      "loss": 0.6077,
+      "step": 23930
+    },
+    {
+      "epoch": 71.89,
+      "grad_norm": 12.077566146850586,
+      "learning_rate": 7.603603603603604e-06,
+      "loss": 0.486,
+      "step": 23940
+    },
+    {
+      "epoch": 71.92,
+      "grad_norm": 17.41840171813965,
+      "learning_rate": 7.6026026026026036e-06,
+      "loss": 0.5488,
+      "step": 23950
+    },
+    {
+      "epoch": 71.95,
+      "grad_norm": 17.22233009338379,
+      "learning_rate": 7.601601601601602e-06,
+      "loss": 0.4889,
+      "step": 23960
+    },
+    {
+      "epoch": 71.98,
+      "grad_norm": 14.471308708190918,
+      "learning_rate": 7.600600600600602e-06,
+      "loss": 0.501,
+      "step": 23970
+    },
+    {
+      "epoch": 72.0,
+      "eval_accuracy": 0.8677,
+      "eval_loss": 0.4600684642791748,
+      "eval_runtime": 12.7849,
+      "eval_samples_per_second": 782.172,
+      "eval_steps_per_second": 3.129,
+      "step": 23976
+    },
+    {
+      "epoch": 72.01,
+      "grad_norm": 15.971799850463867,
+      "learning_rate": 7.599599599599599e-06,
+      "loss": 0.5027,
+      "step": 23980
+    },
+    {
+      "epoch": 72.04,
+      "grad_norm": 15.979533195495605,
+      "learning_rate": 7.598598598598599e-06,
+      "loss": 0.5537,
+      "step": 23990
+    },
+    {
+      "epoch": 72.07,
+      "grad_norm": 16.601274490356445,
+      "learning_rate": 7.597597597597598e-06,
+      "loss": 0.4777,
+      "step": 24000
+    },
+    {
+      "epoch": 72.1,
+      "grad_norm": 13.306760787963867,
+      "learning_rate": 7.596596596596597e-06,
+      "loss": 0.5637,
+      "step": 24010
+    },
+    {
+      "epoch": 72.13,
+      "grad_norm": 19.46541976928711,
+      "learning_rate": 7.595595595595596e-06,
+      "loss": 0.5209,
+      "step": 24020
+    },
+    {
+      "epoch": 72.16,
+      "grad_norm": 10.344480514526367,
+      "learning_rate": 7.594594594594596e-06,
+      "loss": 0.5183,
+      "step": 24030
+    },
+    {
+      "epoch": 72.19,
+      "grad_norm": 15.620344161987305,
+      "learning_rate": 7.593593593593594e-06,
+      "loss": 0.5722,
+      "step": 24040
+    },
+    {
+      "epoch": 72.22,
+      "grad_norm": 16.568490982055664,
+      "learning_rate": 7.592592592592594e-06,
+      "loss": 0.5324,
+      "step": 24050
+    },
+    {
+      "epoch": 72.25,
+      "grad_norm": 11.025808334350586,
+      "learning_rate": 7.591591591591592e-06,
+      "loss": 0.5042,
+      "step": 24060
+    },
+    {
+      "epoch": 72.28,
+      "grad_norm": 14.100255012512207,
+      "learning_rate": 7.590590590590591e-06,
+      "loss": 0.5028,
+      "step": 24070
+    },
+    {
+      "epoch": 72.31,
+      "grad_norm": 11.300607681274414,
+      "learning_rate": 7.5895895895895895e-06,
+      "loss": 0.4996,
+      "step": 24080
+    },
+    {
+      "epoch": 72.34,
+      "grad_norm": 15.547744750976562,
+      "learning_rate": 7.588588588588589e-06,
+      "loss": 0.549,
+      "step": 24090
+    },
+    {
+      "epoch": 72.37,
+      "grad_norm": 17.27249526977539,
+      "learning_rate": 7.587587587587588e-06,
+      "loss": 0.5353,
+      "step": 24100
+    },
+    {
+      "epoch": 72.4,
+      "grad_norm": 11.9943208694458,
+      "learning_rate": 7.586586586586587e-06,
+      "loss": 0.5581,
+      "step": 24110
+    },
+    {
+      "epoch": 72.43,
+      "grad_norm": 9.500299453735352,
+      "learning_rate": 7.5855855855855865e-06,
+      "loss": 0.5421,
+      "step": 24120
+    },
+    {
+      "epoch": 72.46,
+      "grad_norm": 17.034353256225586,
+      "learning_rate": 7.584584584584586e-06,
+      "loss": 0.5442,
+      "step": 24130
+    },
+    {
+      "epoch": 72.49,
+      "grad_norm": 12.58971881866455,
+      "learning_rate": 7.5835835835835845e-06,
+      "loss": 0.5692,
+      "step": 24140
+    },
+    {
+      "epoch": 72.52,
+      "grad_norm": 13.629549980163574,
+      "learning_rate": 7.582582582582583e-06,
+      "loss": 0.561,
+      "step": 24150
+    },
+    {
+      "epoch": 72.55,
+      "grad_norm": 13.327164649963379,
+      "learning_rate": 7.581581581581582e-06,
+      "loss": 0.5717,
+      "step": 24160
+    },
+    {
+      "epoch": 72.58,
+      "grad_norm": 14.426154136657715,
+      "learning_rate": 7.580580580580581e-06,
+      "loss": 0.5369,
+      "step": 24170
+    },
+    {
+      "epoch": 72.61,
+      "grad_norm": 12.07056999206543,
+      "learning_rate": 7.57957957957958e-06,
+      "loss": 0.5526,
+      "step": 24180
+    },
+    {
+      "epoch": 72.64,
+      "grad_norm": 13.597389221191406,
+      "learning_rate": 7.578578578578579e-06,
+      "loss": 0.5784,
+      "step": 24190
+    },
+    {
+      "epoch": 72.67,
+      "grad_norm": 19.362857818603516,
+      "learning_rate": 7.577577577577579e-06,
+      "loss": 0.5699,
+      "step": 24200
+    },
+    {
+      "epoch": 72.7,
+      "grad_norm": 15.222396850585938,
+      "learning_rate": 7.576576576576577e-06,
+      "loss": 0.4642,
+      "step": 24210
+    },
+    {
+      "epoch": 72.73,
+      "grad_norm": 21.673702239990234,
+      "learning_rate": 7.575575575575577e-06,
+      "loss": 0.5819,
+      "step": 24220
+    },
+    {
+      "epoch": 72.76,
+      "grad_norm": 18.147018432617188,
+      "learning_rate": 7.574574574574574e-06,
+      "loss": 0.5674,
+      "step": 24230
+    },
+    {
+      "epoch": 72.79,
+      "grad_norm": 12.631148338317871,
+      "learning_rate": 7.573573573573574e-06,
+      "loss": 0.5157,
+      "step": 24240
+    },
+    {
+      "epoch": 72.82,
+      "grad_norm": 14.27245044708252,
+      "learning_rate": 7.572572572572573e-06,
+      "loss": 0.5214,
+      "step": 24250
+    },
+    {
+      "epoch": 72.85,
+      "grad_norm": 15.539676666259766,
+      "learning_rate": 7.571571571571572e-06,
+      "loss": 0.5118,
+      "step": 24260
+    },
+    {
+      "epoch": 72.88,
+      "grad_norm": 15.721545219421387,
+      "learning_rate": 7.570570570570571e-06,
+      "loss": 0.5546,
+      "step": 24270
+    },
+    {
+      "epoch": 72.91,
+      "grad_norm": 13.978747367858887,
+      "learning_rate": 7.569569569569571e-06,
+      "loss": 0.5183,
+      "step": 24280
+    },
+    {
+      "epoch": 72.94,
+      "grad_norm": 15.935824394226074,
+      "learning_rate": 7.568568568568569e-06,
+      "loss": 0.5509,
+      "step": 24290
+    },
+    {
+      "epoch": 72.97,
+      "grad_norm": 12.331801414489746,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.527,
+      "step": 24300
+    },
+    {
+      "epoch": 73.0,
+      "eval_accuracy": 0.8644,
+      "eval_loss": 0.4660574793815613,
+      "eval_runtime": 12.9139,
+      "eval_samples_per_second": 774.361,
+      "eval_steps_per_second": 3.097,
+      "step": 24309
+    },
+    {
+      "epoch": 73.0,
+      "grad_norm": 15.559767723083496,
+      "learning_rate": 7.566566566566567e-06,
+      "loss": 0.4842,
+      "step": 24310
+    },
+    {
+      "epoch": 73.03,
+      "grad_norm": 16.145465850830078,
+      "learning_rate": 7.565565565565566e-06,
+      "loss": 0.5281,
+      "step": 24320
+    },
+    {
+      "epoch": 73.06,
+      "grad_norm": 14.819173812866211,
+      "learning_rate": 7.5645645645645646e-06,
+      "loss": 0.5776,
+      "step": 24330
+    },
+    {
+      "epoch": 73.09,
+      "grad_norm": 17.449100494384766,
+      "learning_rate": 7.563563563563564e-06,
+      "loss": 0.506,
+      "step": 24340
+    },
+    {
+      "epoch": 73.12,
+      "grad_norm": 17.26336669921875,
+      "learning_rate": 7.5625625625625634e-06,
+      "loss": 0.5094,
+      "step": 24350
+    },
+    {
+      "epoch": 73.15,
+      "grad_norm": 18.2810001373291,
+      "learning_rate": 7.561561561561562e-06,
+      "loss": 0.5599,
+      "step": 24360
+    },
+    {
+      "epoch": 73.18,
+      "grad_norm": 16.956859588623047,
+      "learning_rate": 7.5605605605605615e-06,
+      "loss": 0.5338,
+      "step": 24370
+    },
+    {
+      "epoch": 73.21,
+      "grad_norm": 13.581554412841797,
+      "learning_rate": 7.559559559559561e-06,
+      "loss": 0.5648,
+      "step": 24380
+    },
+    {
+      "epoch": 73.24,
+      "grad_norm": 11.054035186767578,
+      "learning_rate": 7.5585585585585595e-06,
+      "loss": 0.5501,
+      "step": 24390
+    },
+    {
+      "epoch": 73.27,
+      "grad_norm": 17.458660125732422,
+      "learning_rate": 7.557557557557558e-06,
+      "loss": 0.5212,
+      "step": 24400
+    },
+    {
+      "epoch": 73.3,
+      "grad_norm": 11.59261417388916,
+      "learning_rate": 7.556556556556557e-06,
+      "loss": 0.559,
+      "step": 24410
+    },
+    {
+      "epoch": 73.33,
+      "grad_norm": 16.86316680908203,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 0.58,
+      "step": 24420
+    },
+    {
+      "epoch": 73.36,
+      "grad_norm": 19.530414581298828,
+      "learning_rate": 7.554554554554555e-06,
+      "loss": 0.5764,
+      "step": 24430
+    },
+    {
+      "epoch": 73.39,
+      "grad_norm": 12.657683372497559,
+      "learning_rate": 7.553553553553554e-06,
+      "loss": 0.5493,
+      "step": 24440
+    },
+    {
+      "epoch": 73.42,
+      "grad_norm": 14.319578170776367,
+      "learning_rate": 7.552552552552554e-06,
+      "loss": 0.5061,
+      "step": 24450
+    },
+    {
+      "epoch": 73.45,
+      "grad_norm": 15.400496482849121,
+      "learning_rate": 7.551551551551552e-06,
+      "loss": 0.5655,
+      "step": 24460
+    },
+    {
+      "epoch": 73.48,
+      "grad_norm": 20.86901092529297,
+      "learning_rate": 7.550550550550552e-06,
+      "loss": 0.5578,
+      "step": 24470
+    },
+    {
+      "epoch": 73.51,
+      "grad_norm": 15.07816219329834,
+      "learning_rate": 7.549549549549549e-06,
+      "loss": 0.5714,
+      "step": 24480
+    },
+    {
+      "epoch": 73.54,
+      "grad_norm": 17.34552764892578,
+      "learning_rate": 7.548548548548549e-06,
+      "loss": 0.5304,
+      "step": 24490
+    },
+    {
+      "epoch": 73.57,
+      "grad_norm": 14.773884773254395,
+      "learning_rate": 7.547547547547548e-06,
+      "loss": 0.5399,
+      "step": 24500
+    },
+    {
+      "epoch": 73.6,
+      "grad_norm": 19.51259422302246,
+      "learning_rate": 7.546546546546547e-06,
+      "loss": 0.5351,
+      "step": 24510
+    },
+    {
+      "epoch": 73.63,
+      "grad_norm": 16.412368774414062,
+      "learning_rate": 7.545545545545546e-06,
+      "loss": 0.4782,
+      "step": 24520
+    },
+    {
+      "epoch": 73.66,
+      "grad_norm": 15.846576690673828,
+      "learning_rate": 7.544544544544545e-06,
+      "loss": 0.5483,
+      "step": 24530
+    },
+    {
+      "epoch": 73.69,
+      "grad_norm": 15.892062187194824,
+      "learning_rate": 7.543543543543544e-06,
+      "loss": 0.5103,
+      "step": 24540
+    },
+    {
+      "epoch": 73.72,
+      "grad_norm": 14.08358383178711,
+      "learning_rate": 7.542542542542544e-06,
+      "loss": 0.5717,
+      "step": 24550
+    },
+    {
+      "epoch": 73.75,
+      "grad_norm": 14.289830207824707,
+      "learning_rate": 7.5415415415415416e-06,
+      "loss": 0.5413,
+      "step": 24560
+    },
+    {
+      "epoch": 73.78,
+      "grad_norm": 16.19147300720215,
+      "learning_rate": 7.540540540540541e-06,
+      "loss": 0.5163,
+      "step": 24570
+    },
+    {
+      "epoch": 73.81,
+      "grad_norm": 13.138388633728027,
+      "learning_rate": 7.53953953953954e-06,
+      "loss": 0.5051,
+      "step": 24580
+    },
+    {
+      "epoch": 73.84,
+      "grad_norm": 20.523103713989258,
+      "learning_rate": 7.538538538538539e-06,
+      "loss": 0.5301,
+      "step": 24590
+    },
+    {
+      "epoch": 73.87,
+      "grad_norm": 13.263509750366211,
+      "learning_rate": 7.5375375375375385e-06,
+      "loss": 0.5616,
+      "step": 24600
+    },
+    {
+      "epoch": 73.9,
+      "grad_norm": 12.0683012008667,
+      "learning_rate": 7.536536536536537e-06,
+      "loss": 0.5498,
+      "step": 24610
+    },
+    {
+      "epoch": 73.93,
+      "grad_norm": 20.918397903442383,
+      "learning_rate": 7.5355355355355365e-06,
+      "loss": 0.5083,
+      "step": 24620
+    },
+    {
+      "epoch": 73.96,
+      "grad_norm": 20.743820190429688,
+      "learning_rate": 7.534534534534535e-06,
+      "loss": 0.5541,
+      "step": 24630
+    },
+    {
+      "epoch": 73.99,
+      "grad_norm": 26.155380249023438,
+      "learning_rate": 7.5335335335335346e-06,
+      "loss": 0.5618,
+      "step": 24640
+    },
+    {
+      "epoch": 74.0,
+      "eval_accuracy": 0.8664,
+      "eval_loss": 0.46769315004348755,
+      "eval_runtime": 13.034,
+      "eval_samples_per_second": 767.225,
+      "eval_steps_per_second": 3.069,
+      "step": 24642
+    },
+    {
+      "epoch": 74.02,
+      "grad_norm": 14.723753929138184,
+      "learning_rate": 7.532532532532532e-06,
+      "loss": 0.5415,
+      "step": 24650
+    },
+    {
+      "epoch": 74.05,
+      "grad_norm": 14.85951042175293,
+      "learning_rate": 7.531531531531532e-06,
+      "loss": 0.4975,
+      "step": 24660
+    },
+    {
+      "epoch": 74.08,
+      "grad_norm": 15.907417297363281,
+      "learning_rate": 7.530530530530531e-06,
+      "loss": 0.5815,
+      "step": 24670
+    },
+    {
+      "epoch": 74.11,
+      "grad_norm": 18.325946807861328,
+      "learning_rate": 7.52952952952953e-06,
+      "loss": 0.5234,
+      "step": 24680
+    },
+    {
+      "epoch": 74.14,
+      "grad_norm": 13.661330223083496,
+      "learning_rate": 7.528528528528529e-06,
+      "loss": 0.5401,
+      "step": 24690
+    },
+    {
+      "epoch": 74.17,
+      "grad_norm": 17.674766540527344,
+      "learning_rate": 7.527527527527529e-06,
+      "loss": 0.5339,
+      "step": 24700
+    },
+    {
+      "epoch": 74.2,
+      "grad_norm": 14.289420127868652,
+      "learning_rate": 7.526526526526527e-06,
+      "loss": 0.5384,
+      "step": 24710
+    },
+    {
+      "epoch": 74.23,
+      "grad_norm": 15.812458038330078,
+      "learning_rate": 7.525525525525527e-06,
+      "loss": 0.541,
+      "step": 24720
+    },
+    {
+      "epoch": 74.26,
+      "grad_norm": 13.972175598144531,
+      "learning_rate": 7.5245245245245245e-06,
+      "loss": 0.5716,
+      "step": 24730
+    },
+    {
+      "epoch": 74.29,
+      "grad_norm": 14.59525203704834,
+      "learning_rate": 7.523523523523524e-06,
+      "loss": 0.5133,
+      "step": 24740
+    },
+    {
+      "epoch": 74.32,
+      "grad_norm": 15.350430488586426,
+      "learning_rate": 7.5225225225225225e-06,
+      "loss": 0.5353,
+      "step": 24750
+    },
+    {
+      "epoch": 74.35,
+      "grad_norm": 18.458045959472656,
+      "learning_rate": 7.521521521521522e-06,
+      "loss": 0.5476,
+      "step": 24760
+    },
+    {
+      "epoch": 74.38,
+      "grad_norm": 16.9620361328125,
+      "learning_rate": 7.520520520520521e-06,
+      "loss": 0.5185,
+      "step": 24770
+    },
+    {
+      "epoch": 74.41,
+      "grad_norm": 19.114097595214844,
+      "learning_rate": 7.51951951951952e-06,
+      "loss": 0.5245,
+      "step": 24780
+    },
+    {
+      "epoch": 74.44,
+      "grad_norm": 17.223175048828125,
+      "learning_rate": 7.518518518518519e-06,
+      "loss": 0.5483,
+      "step": 24790
+    },
+    {
+      "epoch": 74.47,
+      "grad_norm": 19.7282772064209,
+      "learning_rate": 7.517517517517519e-06,
+      "loss": 0.511,
+      "step": 24800
+    },
+    {
+      "epoch": 74.5,
+      "grad_norm": 13.930578231811523,
+      "learning_rate": 7.516516516516517e-06,
+      "loss": 0.5263,
+      "step": 24810
+    },
+    {
+      "epoch": 74.53,
+      "grad_norm": 12.889962196350098,
+      "learning_rate": 7.515515515515516e-06,
+      "loss": 0.5838,
+      "step": 24820
+    },
+    {
+      "epoch": 74.56,
+      "grad_norm": 13.918925285339355,
+      "learning_rate": 7.514514514514515e-06,
+      "loss": 0.5182,
+      "step": 24830
+    },
+    {
+      "epoch": 74.59,
+      "grad_norm": 13.224120140075684,
+      "learning_rate": 7.513513513513514e-06,
+      "loss": 0.5643,
+      "step": 24840
+    },
+    {
+      "epoch": 74.62,
+      "grad_norm": 13.113409996032715,
+      "learning_rate": 7.5125125125125135e-06,
+      "loss": 0.5046,
+      "step": 24850
+    },
+    {
+      "epoch": 74.65,
+      "grad_norm": 13.158830642700195,
+      "learning_rate": 7.511511511511512e-06,
+      "loss": 0.4659,
+      "step": 24860
+    },
+    {
+      "epoch": 74.68,
+      "grad_norm": 14.555266380310059,
+      "learning_rate": 7.5105105105105116e-06,
+      "loss": 0.4796,
+      "step": 24870
+    },
+    {
+      "epoch": 74.71,
+      "grad_norm": 15.64537239074707,
+      "learning_rate": 7.50950950950951e-06,
+      "loss": 0.5034,
+      "step": 24880
+    },
+    {
+      "epoch": 74.74,
+      "grad_norm": 20.45342254638672,
+      "learning_rate": 7.50850850850851e-06,
+      "loss": 0.5726,
+      "step": 24890
+    },
+    {
+      "epoch": 74.77,
+      "grad_norm": 13.022698402404785,
+      "learning_rate": 7.507507507507507e-06,
+      "loss": 0.5209,
+      "step": 24900
+    },
+    {
+      "epoch": 74.8,
+      "grad_norm": 22.128602981567383,
+      "learning_rate": 7.506506506506507e-06,
+      "loss": 0.6244,
+      "step": 24910
+    },
+    {
+      "epoch": 74.83,
+      "grad_norm": 13.853561401367188,
+      "learning_rate": 7.505505505505506e-06,
+      "loss": 0.5215,
+      "step": 24920
+    },
+    {
+      "epoch": 74.86,
+      "grad_norm": 14.449660301208496,
+      "learning_rate": 7.504504504504505e-06,
+      "loss": 0.5369,
+      "step": 24930
+    },
+    {
+      "epoch": 74.89,
+      "grad_norm": 16.451210021972656,
+      "learning_rate": 7.503503503503504e-06,
+      "loss": 0.537,
+      "step": 24940
+    },
+    {
+      "epoch": 74.92,
+      "grad_norm": 17.204198837280273,
+      "learning_rate": 7.502502502502504e-06,
+      "loss": 0.4918,
+      "step": 24950
+    },
+    {
+      "epoch": 74.95,
+      "grad_norm": 13.26692008972168,
+      "learning_rate": 7.501501501501502e-06,
+      "loss": 0.5672,
+      "step": 24960
+    },
+    {
+      "epoch": 74.98,
+      "grad_norm": 11.932439804077148,
+      "learning_rate": 7.500500500500502e-06,
+      "loss": 0.5161,
+      "step": 24970
+    },
+    {
+      "epoch": 75.0,
+      "eval_accuracy": 0.8691,
+      "eval_loss": 0.4629597067832947,
+      "eval_runtime": 12.3855,
+      "eval_samples_per_second": 807.397,
+      "eval_steps_per_second": 3.23,
+      "step": 24975
+    },
+    {
+      "epoch": 75.02,
+      "grad_norm": 20.42967414855957,
+      "learning_rate": 7.4994994994994995e-06,
+      "loss": 0.4581,
+      "step": 24980
+    },
+    {
+      "epoch": 75.05,
+      "grad_norm": 13.88707160949707,
+      "learning_rate": 7.498498498498499e-06,
+      "loss": 0.4407,
+      "step": 24990
+    },
+    {
+      "epoch": 75.08,
+      "grad_norm": 12.693586349487305,
+      "learning_rate": 7.4974974974974975e-06,
+      "loss": 0.5471,
+      "step": 25000
+    },
+    {
+      "epoch": 75.11,
+      "grad_norm": 18.46637725830078,
+      "learning_rate": 7.496496496496497e-06,
+      "loss": 0.5798,
+      "step": 25010
+    },
+    {
+      "epoch": 75.14,
+      "grad_norm": 22.856239318847656,
+      "learning_rate": 7.495495495495496e-06,
+      "loss": 0.496,
+      "step": 25020
+    },
+    {
+      "epoch": 75.17,
+      "grad_norm": 15.663064002990723,
+      "learning_rate": 7.494494494494495e-06,
+      "loss": 0.5584,
+      "step": 25030
+    },
+    {
+      "epoch": 75.2,
+      "grad_norm": 21.31612205505371,
+      "learning_rate": 7.4934934934934944e-06,
+      "loss": 0.4742,
+      "step": 25040
+    },
+    {
+      "epoch": 75.23,
+      "grad_norm": 13.868929862976074,
+      "learning_rate": 7.492492492492494e-06,
+      "loss": 0.5257,
+      "step": 25050
+    },
+    {
+      "epoch": 75.26,
+      "grad_norm": 13.549492835998535,
+      "learning_rate": 7.491491491491492e-06,
+      "loss": 0.5409,
+      "step": 25060
+    },
+    {
+      "epoch": 75.29,
+      "grad_norm": 13.97533130645752,
+      "learning_rate": 7.490490490490491e-06,
+      "loss": 0.5411,
+      "step": 25070
+    },
+    {
+      "epoch": 75.32,
+      "grad_norm": 14.013882637023926,
+      "learning_rate": 7.48948948948949e-06,
+      "loss": 0.5462,
+      "step": 25080
+    },
+    {
+      "epoch": 75.35,
+      "grad_norm": 16.790821075439453,
+      "learning_rate": 7.488488488488489e-06,
+      "loss": 0.5867,
+      "step": 25090
+    },
+    {
+      "epoch": 75.38,
+      "grad_norm": 20.78099250793457,
+      "learning_rate": 7.487487487487488e-06,
+      "loss": 0.5547,
+      "step": 25100
+    },
+    {
+      "epoch": 75.41,
+      "grad_norm": 11.377620697021484,
+      "learning_rate": 7.486486486486487e-06,
+      "loss": 0.5413,
+      "step": 25110
+    },
+    {
+      "epoch": 75.44,
+      "grad_norm": 27.481689453125,
+      "learning_rate": 7.485485485485487e-06,
+      "loss": 0.5153,
+      "step": 25120
+    },
+    {
+      "epoch": 75.47,
+      "grad_norm": 13.65371036529541,
+      "learning_rate": 7.484484484484485e-06,
+      "loss": 0.5683,
+      "step": 25130
+    },
+    {
+      "epoch": 75.5,
+      "grad_norm": 12.149045944213867,
+      "learning_rate": 7.483483483483485e-06,
+      "loss": 0.5496,
+      "step": 25140
+    },
+    {
+      "epoch": 75.53,
+      "grad_norm": 13.61082935333252,
+      "learning_rate": 7.482482482482482e-06,
+      "loss": 0.4959,
+      "step": 25150
+    },
+    {
+      "epoch": 75.56,
+      "grad_norm": 20.277263641357422,
+      "learning_rate": 7.481481481481482e-06,
+      "loss": 0.4964,
+      "step": 25160
+    },
+    {
+      "epoch": 75.59,
+      "grad_norm": 13.979513168334961,
+      "learning_rate": 7.480480480480481e-06,
+      "loss": 0.5109,
+      "step": 25170
+    },
+    {
+      "epoch": 75.62,
+      "grad_norm": 14.793264389038086,
+      "learning_rate": 7.47947947947948e-06,
+      "loss": 0.5199,
+      "step": 25180
+    },
+    {
+      "epoch": 75.65,
+      "grad_norm": 15.40218448638916,
+      "learning_rate": 7.478478478478479e-06,
+      "loss": 0.517,
+      "step": 25190
+    },
+    {
+      "epoch": 75.68,
+      "grad_norm": 10.543524742126465,
+      "learning_rate": 7.477477477477479e-06,
+      "loss": 0.582,
+      "step": 25200
+    },
+    {
+      "epoch": 75.71,
+      "grad_norm": 28.832815170288086,
+      "learning_rate": 7.476476476476477e-06,
+      "loss": 0.5057,
+      "step": 25210
+    },
+    {
+      "epoch": 75.74,
+      "grad_norm": 12.903115272521973,
+      "learning_rate": 7.475475475475477e-06,
+      "loss": 0.5065,
+      "step": 25220
+    },
+    {
+      "epoch": 75.77,
+      "grad_norm": 24.768508911132812,
+      "learning_rate": 7.4744744744744745e-06,
+      "loss": 0.5163,
+      "step": 25230
+    },
+    {
+      "epoch": 75.8,
+      "grad_norm": 11.10902214050293,
+      "learning_rate": 7.473473473473474e-06,
+      "loss": 0.5206,
+      "step": 25240
+    },
+    {
+      "epoch": 75.83,
+      "grad_norm": 10.880926132202148,
+      "learning_rate": 7.4724724724724726e-06,
+      "loss": 0.49,
+      "step": 25250
+    },
+    {
+      "epoch": 75.86,
+      "grad_norm": 20.442588806152344,
+      "learning_rate": 7.471471471471472e-06,
+      "loss": 0.5405,
+      "step": 25260
+    },
+    {
+      "epoch": 75.89,
+      "grad_norm": 15.126239776611328,
+      "learning_rate": 7.4704704704704714e-06,
+      "loss": 0.4991,
+      "step": 25270
+    },
+    {
+      "epoch": 75.92,
+      "grad_norm": 15.257558822631836,
+      "learning_rate": 7.46946946946947e-06,
+      "loss": 0.5766,
+      "step": 25280
+    },
+    {
+      "epoch": 75.95,
+      "grad_norm": 14.302605628967285,
+      "learning_rate": 7.4684684684684695e-06,
+      "loss": 0.5137,
+      "step": 25290
+    },
+    {
+      "epoch": 75.98,
+      "grad_norm": 15.673317909240723,
+      "learning_rate": 7.467467467467469e-06,
+      "loss": 0.5158,
+      "step": 25300
+    },
+    {
+      "epoch": 76.0,
+      "eval_accuracy": 0.8671,
+      "eval_loss": 0.46911802887916565,
+      "eval_runtime": 12.7971,
+      "eval_samples_per_second": 781.425,
+      "eval_steps_per_second": 3.126,
+      "step": 25308
+    },
+    {
+      "epoch": 76.01,
+      "grad_norm": 14.649646759033203,
+      "learning_rate": 7.466466466466467e-06,
+      "loss": 0.4862,
+      "step": 25310
+    },
+    {
+      "epoch": 76.04,
+      "grad_norm": 13.316688537597656,
+      "learning_rate": 7.465465465465466e-06,
+      "loss": 0.4907,
+      "step": 25320
+    },
+    {
+      "epoch": 76.07,
+      "grad_norm": 18.16513442993164,
+      "learning_rate": 7.464464464464465e-06,
+      "loss": 0.5469,
+      "step": 25330
+    },
+    {
+      "epoch": 76.1,
+      "grad_norm": 18.091474533081055,
+      "learning_rate": 7.463463463463464e-06,
+      "loss": 0.5662,
+      "step": 25340
+    },
+    {
+      "epoch": 76.13,
+      "grad_norm": 13.058123588562012,
+      "learning_rate": 7.462462462462463e-06,
+      "loss": 0.5114,
+      "step": 25350
+    },
+    {
+      "epoch": 76.16,
+      "grad_norm": 32.16606140136719,
+      "learning_rate": 7.461461461461462e-06,
+      "loss": 0.5055,
+      "step": 25360
+    },
+    {
+      "epoch": 76.19,
+      "grad_norm": 10.640914916992188,
+      "learning_rate": 7.460460460460462e-06,
+      "loss": 0.5011,
+      "step": 25370
+    },
+    {
+      "epoch": 76.22,
+      "grad_norm": 12.13187026977539,
+      "learning_rate": 7.45945945945946e-06,
+      "loss": 0.5331,
+      "step": 25380
+    },
+    {
+      "epoch": 76.25,
+      "grad_norm": 22.112607955932617,
+      "learning_rate": 7.45845845845846e-06,
+      "loss": 0.5492,
+      "step": 25390
+    },
+    {
+      "epoch": 76.28,
+      "grad_norm": 20.70840072631836,
+      "learning_rate": 7.457457457457457e-06,
+      "loss": 0.5628,
+      "step": 25400
+    },
+    {
+      "epoch": 76.31,
+      "grad_norm": 19.347606658935547,
+      "learning_rate": 7.456456456456457e-06,
+      "loss": 0.5265,
+      "step": 25410
+    },
+    {
+      "epoch": 76.34,
+      "grad_norm": 15.124089241027832,
+      "learning_rate": 7.455455455455456e-06,
+      "loss": 0.4816,
+      "step": 25420
+    },
+    {
+      "epoch": 76.37,
+      "grad_norm": 11.108996391296387,
+      "learning_rate": 7.454454454454455e-06,
+      "loss": 0.5009,
+      "step": 25430
+    },
+    {
+      "epoch": 76.4,
+      "grad_norm": 24.092477798461914,
+      "learning_rate": 7.453453453453454e-06,
+      "loss": 0.5121,
+      "step": 25440
+    },
+    {
+      "epoch": 76.43,
+      "grad_norm": 17.434934616088867,
+      "learning_rate": 7.452452452452453e-06,
+      "loss": 0.4952,
+      "step": 25450
+    },
+    {
+      "epoch": 76.46,
+      "grad_norm": 18.704214096069336,
+      "learning_rate": 7.451451451451452e-06,
+      "loss": 0.4373,
+      "step": 25460
+    },
+    {
+      "epoch": 76.49,
+      "grad_norm": 18.26543426513672,
+      "learning_rate": 7.450450450450452e-06,
+      "loss": 0.5725,
+      "step": 25470
+    },
+    {
+      "epoch": 76.52,
+      "grad_norm": 25.885238647460938,
+      "learning_rate": 7.4494494494494496e-06,
+      "loss": 0.537,
+      "step": 25480
+    },
+    {
+      "epoch": 76.55,
+      "grad_norm": 30.565876007080078,
+      "learning_rate": 7.448448448448449e-06,
+      "loss": 0.4964,
+      "step": 25490
+    },
+    {
+      "epoch": 76.58,
+      "grad_norm": 20.226112365722656,
+      "learning_rate": 7.447447447447448e-06,
+      "loss": 0.5429,
+      "step": 25500
+    },
+    {
+      "epoch": 76.61,
+      "grad_norm": 18.410228729248047,
+      "learning_rate": 7.446446446446447e-06,
+      "loss": 0.4911,
+      "step": 25510
+    },
+    {
+      "epoch": 76.64,
+      "grad_norm": 24.221858978271484,
+      "learning_rate": 7.4454454454454465e-06,
+      "loss": 0.4685,
+      "step": 25520
+    },
+    {
+      "epoch": 76.67,
+      "grad_norm": 20.43824577331543,
+      "learning_rate": 7.444444444444445e-06,
+      "loss": 0.5016,
+      "step": 25530
+    },
+    {
+      "epoch": 76.7,
+      "grad_norm": 13.232439994812012,
+      "learning_rate": 7.4434434434434445e-06,
+      "loss": 0.5267,
+      "step": 25540
+    },
+    {
+      "epoch": 76.73,
+      "grad_norm": 16.06315040588379,
+      "learning_rate": 7.442442442442443e-06,
+      "loss": 0.5382,
+      "step": 25550
+    },
+    {
+      "epoch": 76.76,
+      "grad_norm": 15.746918678283691,
+      "learning_rate": 7.441441441441442e-06,
+      "loss": 0.5118,
+      "step": 25560
+    },
+    {
+      "epoch": 76.79,
+      "grad_norm": 22.798219680786133,
+      "learning_rate": 7.44044044044044e-06,
+      "loss": 0.4976,
+      "step": 25570
+    },
+    {
+      "epoch": 76.82,
+      "grad_norm": 15.366040229797363,
+      "learning_rate": 7.43943943943944e-06,
+      "loss": 0.5226,
+      "step": 25580
+    },
+    {
+      "epoch": 76.85,
+      "grad_norm": 15.96611499786377,
+      "learning_rate": 7.438438438438439e-06,
+      "loss": 0.5613,
+      "step": 25590
+    },
+    {
+      "epoch": 76.88,
+      "grad_norm": 19.329378128051758,
+      "learning_rate": 7.437437437437438e-06,
+      "loss": 0.4833,
+      "step": 25600
+    },
+    {
+      "epoch": 76.91,
+      "grad_norm": 18.379074096679688,
+      "learning_rate": 7.436436436436437e-06,
+      "loss": 0.5026,
+      "step": 25610
+    },
+    {
+      "epoch": 76.94,
+      "grad_norm": 12.549742698669434,
+      "learning_rate": 7.435435435435437e-06,
+      "loss": 0.5504,
+      "step": 25620
+    },
+    {
+      "epoch": 76.97,
+      "grad_norm": 15.263510704040527,
+      "learning_rate": 7.434434434434435e-06,
+      "loss": 0.4608,
+      "step": 25630
+    },
+    {
+      "epoch": 77.0,
+      "grad_norm": 21.36216163635254,
+      "learning_rate": 7.433433433433434e-06,
+      "loss": 0.54,
+      "step": 25640
+    },
+    {
+      "epoch": 77.0,
+      "eval_accuracy": 0.8696,
+      "eval_loss": 0.46450892090797424,
+      "eval_runtime": 13.007,
+      "eval_samples_per_second": 768.818,
+      "eval_steps_per_second": 3.075,
+      "step": 25641
+    },
+    {
+      "epoch": 77.03,
+      "grad_norm": 14.115147590637207,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 0.4776,
+      "step": 25650
+    },
+    {
+      "epoch": 77.06,
+      "grad_norm": 19.314167022705078,
+      "learning_rate": 7.431431431431432e-06,
+      "loss": 0.5811,
+      "step": 25660
+    },
+    {
+      "epoch": 77.09,
+      "grad_norm": 11.57456111907959,
+      "learning_rate": 7.4304304304304305e-06,
+      "loss": 0.5489,
+      "step": 25670
+    },
+    {
+      "epoch": 77.12,
+      "grad_norm": 17.31260108947754,
+      "learning_rate": 7.42942942942943e-06,
+      "loss": 0.5489,
+      "step": 25680
+    },
+    {
+      "epoch": 77.15,
+      "grad_norm": 18.1948299407959,
+      "learning_rate": 7.428428428428429e-06,
+      "loss": 0.4979,
+      "step": 25690
+    },
+    {
+      "epoch": 77.18,
+      "grad_norm": 14.55479907989502,
+      "learning_rate": 7.427427427427428e-06,
+      "loss": 0.5327,
+      "step": 25700
+    },
+    {
+      "epoch": 77.21,
+      "grad_norm": 24.24778938293457,
+      "learning_rate": 7.426426426426427e-06,
+      "loss": 0.4662,
+      "step": 25710
+    },
+    {
+      "epoch": 77.24,
+      "grad_norm": 13.166237831115723,
+      "learning_rate": 7.425425425425427e-06,
+      "loss": 0.4897,
+      "step": 25720
+    },
+    {
+      "epoch": 77.27,
+      "grad_norm": 14.406869888305664,
+      "learning_rate": 7.424424424424425e-06,
+      "loss": 0.4897,
+      "step": 25730
+    },
+    {
+      "epoch": 77.3,
+      "grad_norm": 16.864351272583008,
+      "learning_rate": 7.423423423423424e-06,
+      "loss": 0.5531,
+      "step": 25740
+    },
+    {
+      "epoch": 77.33,
+      "grad_norm": 17.601715087890625,
+      "learning_rate": 7.422422422422423e-06,
+      "loss": 0.4555,
+      "step": 25750
+    },
+    {
+      "epoch": 77.36,
+      "grad_norm": 12.461166381835938,
+      "learning_rate": 7.421421421421422e-06,
+      "loss": 0.4522,
+      "step": 25760
+    },
+    {
+      "epoch": 77.39,
+      "grad_norm": 20.531606674194336,
+      "learning_rate": 7.4204204204204215e-06,
+      "loss": 0.544,
+      "step": 25770
+    },
+    {
+      "epoch": 77.42,
+      "grad_norm": 14.070215225219727,
+      "learning_rate": 7.41941941941942e-06,
+      "loss": 0.5131,
+      "step": 25780
+    },
+    {
+      "epoch": 77.45,
+      "grad_norm": 14.558828353881836,
+      "learning_rate": 7.4184184184184195e-06,
+      "loss": 0.5136,
+      "step": 25790
+    },
+    {
+      "epoch": 77.48,
+      "grad_norm": 19.045860290527344,
+      "learning_rate": 7.417417417417418e-06,
+      "loss": 0.5237,
+      "step": 25800
+    },
+    {
+      "epoch": 77.51,
+      "grad_norm": 24.205039978027344,
+      "learning_rate": 7.416416416416417e-06,
+      "loss": 0.5299,
+      "step": 25810
+    },
+    {
+      "epoch": 77.54,
+      "grad_norm": 12.875337600708008,
+      "learning_rate": 7.415415415415415e-06,
+      "loss": 0.5001,
+      "step": 25820
+    },
+    {
+      "epoch": 77.57,
+      "grad_norm": 19.669151306152344,
+      "learning_rate": 7.414414414414415e-06,
+      "loss": 0.5513,
+      "step": 25830
+    },
+    {
+      "epoch": 77.6,
+      "grad_norm": 18.091136932373047,
+      "learning_rate": 7.413413413413414e-06,
+      "loss": 0.5318,
+      "step": 25840
+    },
+    {
+      "epoch": 77.63,
+      "grad_norm": 13.371623039245605,
+      "learning_rate": 7.412412412412413e-06,
+      "loss": 0.5119,
+      "step": 25850
+    },
+    {
+      "epoch": 77.66,
+      "grad_norm": 17.000181198120117,
+      "learning_rate": 7.411411411411412e-06,
+      "loss": 0.5288,
+      "step": 25860
+    },
+    {
+      "epoch": 77.69,
+      "grad_norm": 19.7879638671875,
+      "learning_rate": 7.410410410410412e-06,
+      "loss": 0.555,
+      "step": 25870
+    },
+    {
+      "epoch": 77.72,
+      "grad_norm": 18.344942092895508,
+      "learning_rate": 7.40940940940941e-06,
+      "loss": 0.5115,
+      "step": 25880
+    },
+    {
+      "epoch": 77.75,
+      "grad_norm": 15.039511680603027,
+      "learning_rate": 7.408408408408409e-06,
+      "loss": 0.5399,
+      "step": 25890
+    },
+    {
+      "epoch": 77.78,
+      "grad_norm": 12.591846466064453,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.4651,
+      "step": 25900
+    },
+    {
+      "epoch": 77.81,
+      "grad_norm": 15.273913383483887,
+      "learning_rate": 7.406406406406407e-06,
+      "loss": 0.5634,
+      "step": 25910
+    },
+    {
+      "epoch": 77.84,
+      "grad_norm": 14.827644348144531,
+      "learning_rate": 7.4054054054054055e-06,
+      "loss": 0.4924,
+      "step": 25920
+    },
+    {
+      "epoch": 77.87,
+      "grad_norm": 17.303077697753906,
+      "learning_rate": 7.404404404404405e-06,
+      "loss": 0.5312,
+      "step": 25930
+    },
+    {
+      "epoch": 77.9,
+      "grad_norm": 17.895198822021484,
+      "learning_rate": 7.403403403403404e-06,
+      "loss": 0.5065,
+      "step": 25940
+    },
+    {
+      "epoch": 77.93,
+      "grad_norm": 14.915367126464844,
+      "learning_rate": 7.402402402402403e-06,
+      "loss": 0.5569,
+      "step": 25950
+    },
+    {
+      "epoch": 77.96,
+      "grad_norm": 20.017568588256836,
+      "learning_rate": 7.4014014014014024e-06,
+      "loss": 0.5149,
+      "step": 25960
+    },
+    {
+      "epoch": 77.99,
+      "grad_norm": 13.076175689697266,
+      "learning_rate": 7.400400400400402e-06,
+      "loss": 0.5352,
+      "step": 25970
+    },
+    {
+      "epoch": 78.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.48047134280204773,
+      "eval_runtime": 12.8629,
+      "eval_samples_per_second": 777.431,
+      "eval_steps_per_second": 3.11,
+      "step": 25974
+    },
+    {
+      "epoch": 78.02,
+      "grad_norm": 12.429010391235352,
+      "learning_rate": 7.3993993993994e-06,
+      "loss": 0.48,
+      "step": 25980
+    },
+    {
+      "epoch": 78.05,
+      "grad_norm": 12.912761688232422,
+      "learning_rate": 7.398398398398399e-06,
+      "loss": 0.5868,
+      "step": 25990
+    },
+    {
+      "epoch": 78.08,
+      "grad_norm": 12.608386039733887,
+      "learning_rate": 7.397397397397398e-06,
+      "loss": 0.5076,
+      "step": 26000
+    },
+    {
+      "epoch": 78.11,
+      "grad_norm": 17.142749786376953,
+      "learning_rate": 7.396396396396397e-06,
+      "loss": 0.5145,
+      "step": 26010
+    },
+    {
+      "epoch": 78.14,
+      "grad_norm": 18.803665161132812,
+      "learning_rate": 7.395395395395396e-06,
+      "loss": 0.4547,
+      "step": 26020
+    },
+    {
+      "epoch": 78.17,
+      "grad_norm": 18.04738426208496,
+      "learning_rate": 7.394394394394395e-06,
+      "loss": 0.5698,
+      "step": 26030
+    },
+    {
+      "epoch": 78.2,
+      "grad_norm": 15.603890419006348,
+      "learning_rate": 7.393393393393395e-06,
+      "loss": 0.5145,
+      "step": 26040
+    },
+    {
+      "epoch": 78.23,
+      "grad_norm": 13.2296142578125,
+      "learning_rate": 7.392392392392393e-06,
+      "loss": 0.487,
+      "step": 26050
+    },
+    {
+      "epoch": 78.26,
+      "grad_norm": 20.099300384521484,
+      "learning_rate": 7.391391391391392e-06,
+      "loss": 0.524,
+      "step": 26060
+    },
+    {
+      "epoch": 78.29,
+      "grad_norm": 12.897528648376465,
+      "learning_rate": 7.39039039039039e-06,
+      "loss": 0.5418,
+      "step": 26070
+    },
+    {
+      "epoch": 78.32,
+      "grad_norm": 15.419456481933594,
+      "learning_rate": 7.38938938938939e-06,
+      "loss": 0.555,
+      "step": 26080
+    },
+    {
+      "epoch": 78.35,
+      "grad_norm": 16.353063583374023,
+      "learning_rate": 7.388388388388389e-06,
+      "loss": 0.5425,
+      "step": 26090
+    },
+    {
+      "epoch": 78.38,
+      "grad_norm": 15.16103458404541,
+      "learning_rate": 7.387387387387388e-06,
+      "loss": 0.526,
+      "step": 26100
+    },
+    {
+      "epoch": 78.41,
+      "grad_norm": 12.94898509979248,
+      "learning_rate": 7.386386386386387e-06,
+      "loss": 0.5487,
+      "step": 26110
+    },
+    {
+      "epoch": 78.44,
+      "grad_norm": 17.046621322631836,
+      "learning_rate": 7.385385385385386e-06,
+      "loss": 0.5144,
+      "step": 26120
+    },
+    {
+      "epoch": 78.47,
+      "grad_norm": 19.95154571533203,
+      "learning_rate": 7.384384384384385e-06,
+      "loss": 0.5222,
+      "step": 26130
+    },
+    {
+      "epoch": 78.5,
+      "grad_norm": 18.736618041992188,
+      "learning_rate": 7.383383383383383e-06,
+      "loss": 0.5622,
+      "step": 26140
+    },
+    {
+      "epoch": 78.53,
+      "grad_norm": 18.860132217407227,
+      "learning_rate": 7.3823823823823825e-06,
+      "loss": 0.5259,
+      "step": 26150
+    },
+    {
+      "epoch": 78.56,
+      "grad_norm": 17.85606575012207,
+      "learning_rate": 7.381381381381382e-06,
+      "loss": 0.4899,
+      "step": 26160
+    },
+    {
+      "epoch": 78.59,
+      "grad_norm": 13.834342002868652,
+      "learning_rate": 7.3803803803803806e-06,
+      "loss": 0.4883,
+      "step": 26170
+    },
+    {
+      "epoch": 78.62,
+      "grad_norm": 17.14006233215332,
+      "learning_rate": 7.37937937937938e-06,
+      "loss": 0.5362,
+      "step": 26180
+    },
+    {
+      "epoch": 78.65,
+      "grad_norm": 10.08365535736084,
+      "learning_rate": 7.3783783783783794e-06,
+      "loss": 0.4879,
+      "step": 26190
+    },
+    {
+      "epoch": 78.68,
+      "grad_norm": 13.915678977966309,
+      "learning_rate": 7.377377377377378e-06,
+      "loss": 0.5099,
+      "step": 26200
+    },
+    {
+      "epoch": 78.71,
+      "grad_norm": 12.106389045715332,
+      "learning_rate": 7.3763763763763775e-06,
+      "loss": 0.508,
+      "step": 26210
+    },
+    {
+      "epoch": 78.74,
+      "grad_norm": 21.481420516967773,
+      "learning_rate": 7.375375375375377e-06,
+      "loss": 0.5478,
+      "step": 26220
+    },
+    {
+      "epoch": 78.77,
+      "grad_norm": 10.372005462646484,
+      "learning_rate": 7.374374374374375e-06,
+      "loss": 0.5376,
+      "step": 26230
+    },
+    {
+      "epoch": 78.8,
+      "grad_norm": 11.340289115905762,
+      "learning_rate": 7.373373373373373e-06,
+      "loss": 0.5383,
+      "step": 26240
+    },
+    {
+      "epoch": 78.83,
+      "grad_norm": 11.110851287841797,
+      "learning_rate": 7.372372372372373e-06,
+      "loss": 0.5323,
+      "step": 26250
+    },
+    {
+      "epoch": 78.86,
+      "grad_norm": 20.465383529663086,
+      "learning_rate": 7.371371371371372e-06,
+      "loss": 0.5645,
+      "step": 26260
+    },
+    {
+      "epoch": 78.89,
+      "grad_norm": 14.545747756958008,
+      "learning_rate": 7.370370370370371e-06,
+      "loss": 0.5126,
+      "step": 26270
+    },
+    {
+      "epoch": 78.92,
+      "grad_norm": 16.517396926879883,
+      "learning_rate": 7.36936936936937e-06,
+      "loss": 0.5487,
+      "step": 26280
+    },
+    {
+      "epoch": 78.95,
+      "grad_norm": 17.911067962646484,
+      "learning_rate": 7.36836836836837e-06,
+      "loss": 0.4834,
+      "step": 26290
+    },
+    {
+      "epoch": 78.98,
+      "grad_norm": 12.054357528686523,
+      "learning_rate": 7.367367367367368e-06,
+      "loss": 0.5433,
+      "step": 26300
+    },
+    {
+      "epoch": 79.0,
+      "eval_accuracy": 0.867,
+      "eval_loss": 0.4695671498775482,
+      "eval_runtime": 12.7001,
+      "eval_samples_per_second": 787.398,
+      "eval_steps_per_second": 3.15,
+      "step": 26307
+    },
+    {
+      "epoch": 79.01,
+      "grad_norm": 16.02682113647461,
+      "learning_rate": 7.366366366366367e-06,
+      "loss": 0.5008,
+      "step": 26310
+    },
+    {
+      "epoch": 79.04,
+      "grad_norm": 8.246256828308105,
+      "learning_rate": 7.365365365365365e-06,
+      "loss": 0.4939,
+      "step": 26320
+    },
+    {
+      "epoch": 79.07,
+      "grad_norm": 24.42353630065918,
+      "learning_rate": 7.364364364364365e-06,
+      "loss": 0.5352,
+      "step": 26330
+    },
+    {
+      "epoch": 79.1,
+      "grad_norm": 18.5328369140625,
+      "learning_rate": 7.363363363363364e-06,
+      "loss": 0.5155,
+      "step": 26340
+    },
+    {
+      "epoch": 79.13,
+      "grad_norm": 15.051130294799805,
+      "learning_rate": 7.362362362362363e-06,
+      "loss": 0.4789,
+      "step": 26350
+    },
+    {
+      "epoch": 79.16,
+      "grad_norm": 18.929536819458008,
+      "learning_rate": 7.361361361361362e-06,
+      "loss": 0.572,
+      "step": 26360
+    },
+    {
+      "epoch": 79.19,
+      "grad_norm": 17.490177154541016,
+      "learning_rate": 7.360360360360361e-06,
+      "loss": 0.4708,
+      "step": 26370
+    },
+    {
+      "epoch": 79.22,
+      "grad_norm": 17.31424331665039,
+      "learning_rate": 7.35935935935936e-06,
+      "loss": 0.49,
+      "step": 26380
+    },
+    {
+      "epoch": 79.25,
+      "grad_norm": 18.38802719116211,
+      "learning_rate": 7.358358358358358e-06,
+      "loss": 0.4939,
+      "step": 26390
+    },
+    {
+      "epoch": 79.28,
+      "grad_norm": 15.984980583190918,
+      "learning_rate": 7.3573573573573575e-06,
+      "loss": 0.5251,
+      "step": 26400
+    },
+    {
+      "epoch": 79.31,
+      "grad_norm": 17.431676864624023,
+      "learning_rate": 7.356356356356357e-06,
+      "loss": 0.4669,
+      "step": 26410
+    },
+    {
+      "epoch": 79.34,
+      "grad_norm": 15.049677848815918,
+      "learning_rate": 7.355355355355356e-06,
+      "loss": 0.5497,
+      "step": 26420
+    },
+    {
+      "epoch": 79.37,
+      "grad_norm": 19.36207389831543,
+      "learning_rate": 7.354354354354355e-06,
+      "loss": 0.5202,
+      "step": 26430
+    },
+    {
+      "epoch": 79.4,
+      "grad_norm": 19.58750343322754,
+      "learning_rate": 7.3533533533533545e-06,
+      "loss": 0.5266,
+      "step": 26440
+    },
+    {
+      "epoch": 79.43,
+      "grad_norm": 25.741657257080078,
+      "learning_rate": 7.352352352352353e-06,
+      "loss": 0.5046,
+      "step": 26450
+    },
+    {
+      "epoch": 79.46,
+      "grad_norm": 13.647472381591797,
+      "learning_rate": 7.3513513513513525e-06,
+      "loss": 0.4464,
+      "step": 26460
+    },
+    {
+      "epoch": 79.49,
+      "grad_norm": 13.559154510498047,
+      "learning_rate": 7.350350350350351e-06,
+      "loss": 0.5424,
+      "step": 26470
+    },
+    {
+      "epoch": 79.52,
+      "grad_norm": 11.44558334350586,
+      "learning_rate": 7.34934934934935e-06,
+      "loss": 0.4915,
+      "step": 26480
+    },
+    {
+      "epoch": 79.55,
+      "grad_norm": 13.8262939453125,
+      "learning_rate": 7.348348348348348e-06,
+      "loss": 0.4972,
+      "step": 26490
+    },
+    {
+      "epoch": 79.58,
+      "grad_norm": 12.904470443725586,
+      "learning_rate": 7.347347347347348e-06,
+      "loss": 0.5266,
+      "step": 26500
+    },
+    {
+      "epoch": 79.61,
+      "grad_norm": 15.852788925170898,
+      "learning_rate": 7.346346346346347e-06,
+      "loss": 0.4829,
+      "step": 26510
+    },
+    {
+      "epoch": 79.64,
+      "grad_norm": 12.143349647521973,
+      "learning_rate": 7.345345345345346e-06,
+      "loss": 0.5313,
+      "step": 26520
+    },
+    {
+      "epoch": 79.67,
+      "grad_norm": 16.085865020751953,
+      "learning_rate": 7.344344344344345e-06,
+      "loss": 0.4879,
+      "step": 26530
+    },
+    {
+      "epoch": 79.7,
+      "grad_norm": 16.444753646850586,
+      "learning_rate": 7.343343343343345e-06,
+      "loss": 0.5562,
+      "step": 26540
+    },
+    {
+      "epoch": 79.73,
+      "grad_norm": 15.056065559387207,
+      "learning_rate": 7.342342342342343e-06,
+      "loss": 0.5133,
+      "step": 26550
+    },
+    {
+      "epoch": 79.76,
+      "grad_norm": 21.398672103881836,
+      "learning_rate": 7.341341341341342e-06,
+      "loss": 0.4803,
+      "step": 26560
+    },
+    {
+      "epoch": 79.79,
+      "grad_norm": 17.769912719726562,
+      "learning_rate": 7.3403403403403404e-06,
+      "loss": 0.5133,
+      "step": 26570
+    },
+    {
+      "epoch": 79.82,
+      "grad_norm": 16.689434051513672,
+      "learning_rate": 7.33933933933934e-06,
+      "loss": 0.5437,
+      "step": 26580
+    },
+    {
+      "epoch": 79.85,
+      "grad_norm": 16.95115089416504,
+      "learning_rate": 7.3383383383383385e-06,
+      "loss": 0.536,
+      "step": 26590
+    },
+    {
+      "epoch": 79.88,
+      "grad_norm": 20.207969665527344,
+      "learning_rate": 7.337337337337338e-06,
+      "loss": 0.5318,
+      "step": 26600
+    },
+    {
+      "epoch": 79.91,
+      "grad_norm": 19.271556854248047,
+      "learning_rate": 7.336336336336337e-06,
+      "loss": 0.5122,
+      "step": 26610
+    },
+    {
+      "epoch": 79.94,
+      "grad_norm": 15.942670822143555,
+      "learning_rate": 7.335335335335336e-06,
+      "loss": 0.513,
+      "step": 26620
+    },
+    {
+      "epoch": 79.97,
+      "grad_norm": 18.47221565246582,
+      "learning_rate": 7.334334334334335e-06,
+      "loss": 0.5029,
+      "step": 26630
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 129.9085235595703,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.5555,
+      "step": 26640
+    },
+    {
+      "epoch": 80.0,
+      "eval_accuracy": 0.8657,
+      "eval_loss": 0.4745276868343353,
+      "eval_runtime": 13.0471,
+      "eval_samples_per_second": 766.451,
+      "eval_steps_per_second": 3.066,
+      "step": 26640
+    },
+    {
+      "epoch": 80.03,
+      "grad_norm": 19.516761779785156,
+      "learning_rate": 7.332332332332333e-06,
+      "loss": 0.5288,
+      "step": 26650
+    },
+    {
+      "epoch": 80.06,
+      "grad_norm": 15.747294425964355,
+      "learning_rate": 7.331331331331332e-06,
+      "loss": 0.495,
+      "step": 26660
+    },
+    {
+      "epoch": 80.09,
+      "grad_norm": 15.504973411560059,
+      "learning_rate": 7.330330330330331e-06,
+      "loss": 0.5094,
+      "step": 26670
+    },
+    {
+      "epoch": 80.12,
+      "grad_norm": 11.713431358337402,
+      "learning_rate": 7.32932932932933e-06,
+      "loss": 0.5242,
+      "step": 26680
+    },
+    {
+      "epoch": 80.15,
+      "grad_norm": 14.051051139831543,
+      "learning_rate": 7.328328328328329e-06,
+      "loss": 0.5044,
+      "step": 26690
+    },
+    {
+      "epoch": 80.18,
+      "grad_norm": 16.7624568939209,
+      "learning_rate": 7.327327327327328e-06,
+      "loss": 0.536,
+      "step": 26700
+    },
+    {
+      "epoch": 80.21,
+      "grad_norm": 12.7097806930542,
+      "learning_rate": 7.3263263263263275e-06,
+      "loss": 0.529,
+      "step": 26710
+    },
+    {
+      "epoch": 80.24,
+      "grad_norm": 17.222837448120117,
+      "learning_rate": 7.325325325325326e-06,
+      "loss": 0.52,
+      "step": 26720
+    },
+    {
+      "epoch": 80.27,
+      "grad_norm": 14.705011367797852,
+      "learning_rate": 7.324324324324325e-06,
+      "loss": 0.5685,
+      "step": 26730
+    },
+    {
+      "epoch": 80.3,
+      "grad_norm": 17.78485870361328,
+      "learning_rate": 7.323323323323323e-06,
+      "loss": 0.5046,
+      "step": 26740
+    },
+    {
+      "epoch": 80.33,
+      "grad_norm": 12.968180656433105,
+      "learning_rate": 7.322322322322323e-06,
+      "loss": 0.5051,
+      "step": 26750
+    },
+    {
+      "epoch": 80.36,
+      "grad_norm": 20.615371704101562,
+      "learning_rate": 7.321321321321322e-06,
+      "loss": 0.5554,
+      "step": 26760
+    },
+    {
+      "epoch": 80.39,
+      "grad_norm": 20.71367073059082,
+      "learning_rate": 7.320320320320321e-06,
+      "loss": 0.4645,
+      "step": 26770
+    },
+    {
+      "epoch": 80.42,
+      "grad_norm": 14.724589347839355,
+      "learning_rate": 7.31931931931932e-06,
+      "loss": 0.5227,
+      "step": 26780
+    },
+    {
+      "epoch": 80.45,
+      "grad_norm": 19.262706756591797,
+      "learning_rate": 7.31831831831832e-06,
+      "loss": 0.5481,
+      "step": 26790
+    },
+    {
+      "epoch": 80.48,
+      "grad_norm": 13.716856002807617,
+      "learning_rate": 7.317317317317318e-06,
+      "loss": 0.4711,
+      "step": 26800
+    },
+    {
+      "epoch": 80.51,
+      "grad_norm": 17.449464797973633,
+      "learning_rate": 7.316316316316316e-06,
+      "loss": 0.4541,
+      "step": 26810
+    },
+    {
+      "epoch": 80.54,
+      "grad_norm": 16.110658645629883,
+      "learning_rate": 7.3153153153153155e-06,
+      "loss": 0.5067,
+      "step": 26820
+    },
+    {
+      "epoch": 80.57,
+      "grad_norm": 9.302728652954102,
+      "learning_rate": 7.314314314314315e-06,
+      "loss": 0.5089,
+      "step": 26830
+    },
+    {
+      "epoch": 80.6,
+      "grad_norm": 16.48388671875,
+      "learning_rate": 7.3133133133133135e-06,
+      "loss": 0.5377,
+      "step": 26840
+    },
+    {
+      "epoch": 80.63,
+      "grad_norm": 13.761683464050293,
+      "learning_rate": 7.312312312312313e-06,
+      "loss": 0.4673,
+      "step": 26850
+    },
+    {
+      "epoch": 80.66,
+      "grad_norm": 16.251955032348633,
+      "learning_rate": 7.311311311311312e-06,
+      "loss": 0.5625,
+      "step": 26860
+    },
+    {
+      "epoch": 80.69,
+      "grad_norm": 21.661922454833984,
+      "learning_rate": 7.310310310310311e-06,
+      "loss": 0.536,
+      "step": 26870
+    },
+    {
+      "epoch": 80.72,
+      "grad_norm": 18.71791648864746,
+      "learning_rate": 7.3093093093093104e-06,
+      "loss": 0.5402,
+      "step": 26880
+    },
+    {
+      "epoch": 80.75,
+      "grad_norm": 13.311158180236816,
+      "learning_rate": 7.308308308308308e-06,
+      "loss": 0.4404,
+      "step": 26890
+    },
+    {
+      "epoch": 80.78,
+      "grad_norm": 14.995808601379395,
+      "learning_rate": 7.307307307307308e-06,
+      "loss": 0.5359,
+      "step": 26900
+    },
+    {
+      "epoch": 80.81,
+      "grad_norm": 17.181751251220703,
+      "learning_rate": 7.306306306306307e-06,
+      "loss": 0.5081,
+      "step": 26910
+    },
+    {
+      "epoch": 80.84,
+      "grad_norm": 12.407795906066895,
+      "learning_rate": 7.305305305305306e-06,
+      "loss": 0.5269,
+      "step": 26920
+    },
+    {
+      "epoch": 80.87,
+      "grad_norm": 13.916088104248047,
+      "learning_rate": 7.304304304304305e-06,
+      "loss": 0.5152,
+      "step": 26930
+    },
+    {
+      "epoch": 80.9,
+      "grad_norm": 23.82271957397461,
+      "learning_rate": 7.303303303303304e-06,
+      "loss": 0.5478,
+      "step": 26940
+    },
+    {
+      "epoch": 80.93,
+      "grad_norm": 10.362595558166504,
+      "learning_rate": 7.302302302302303e-06,
+      "loss": 0.4607,
+      "step": 26950
+    },
+    {
+      "epoch": 80.96,
+      "grad_norm": 13.769407272338867,
+      "learning_rate": 7.3013013013013026e-06,
+      "loss": 0.453,
+      "step": 26960
+    },
+    {
+      "epoch": 80.99,
+      "grad_norm": 17.067766189575195,
+      "learning_rate": 7.3003003003003e-06,
+      "loss": 0.5248,
+      "step": 26970
+    },
+    {
+      "epoch": 81.0,
+      "eval_accuracy": 0.8655,
+      "eval_loss": 0.47673192620277405,
+      "eval_runtime": 12.8835,
+      "eval_samples_per_second": 776.186,
+      "eval_steps_per_second": 3.105,
+      "step": 26973
+    },
+    {
+      "epoch": 81.02,
+      "grad_norm": 19.173913955688477,
+      "learning_rate": 7.2992992992993e-06,
+      "loss": 0.5459,
+      "step": 26980
+    },
+    {
+      "epoch": 81.05,
+      "grad_norm": 17.089609146118164,
+      "learning_rate": 7.298298298298298e-06,
+      "loss": 0.469,
+      "step": 26990
+    },
+    {
+      "epoch": 81.08,
+      "grad_norm": 19.796131134033203,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.5065,
+      "step": 27000
+    },
+    {
+      "epoch": 81.11,
+      "grad_norm": 13.776817321777344,
+      "learning_rate": 7.296296296296297e-06,
+      "loss": 0.5056,
+      "step": 27010
+    },
+    {
+      "epoch": 81.14,
+      "grad_norm": 17.408824920654297,
+      "learning_rate": 7.295295295295296e-06,
+      "loss": 0.511,
+      "step": 27020
+    },
+    {
+      "epoch": 81.17,
+      "grad_norm": 41.608097076416016,
+      "learning_rate": 7.294294294294295e-06,
+      "loss": 0.5315,
+      "step": 27030
+    },
+    {
+      "epoch": 81.2,
+      "grad_norm": 14.133171081542969,
+      "learning_rate": 7.293293293293294e-06,
+      "loss": 0.4975,
+      "step": 27040
+    },
+    {
+      "epoch": 81.23,
+      "grad_norm": 14.316442489624023,
+      "learning_rate": 7.292292292292293e-06,
+      "loss": 0.5426,
+      "step": 27050
+    },
+    {
+      "epoch": 81.26,
+      "grad_norm": 15.659419059753418,
+      "learning_rate": 7.291291291291291e-06,
+      "loss": 0.4493,
+      "step": 27060
+    },
+    {
+      "epoch": 81.29,
+      "grad_norm": 16.83542823791504,
+      "learning_rate": 7.2902902902902905e-06,
+      "loss": 0.5195,
+      "step": 27070
+    },
+    {
+      "epoch": 81.32,
+      "grad_norm": 15.329632759094238,
+      "learning_rate": 7.28928928928929e-06,
+      "loss": 0.511,
+      "step": 27080
+    },
+    {
+      "epoch": 81.35,
+      "grad_norm": 15.858656883239746,
+      "learning_rate": 7.2882882882882885e-06,
+      "loss": 0.4848,
+      "step": 27090
+    },
+    {
+      "epoch": 81.38,
+      "grad_norm": 17.962413787841797,
+      "learning_rate": 7.287287287287288e-06,
+      "loss": 0.4607,
+      "step": 27100
+    },
+    {
+      "epoch": 81.41,
+      "grad_norm": 13.855925559997559,
+      "learning_rate": 7.2862862862862874e-06,
+      "loss": 0.4947,
+      "step": 27110
+    },
+    {
+      "epoch": 81.44,
+      "grad_norm": 15.741239547729492,
+      "learning_rate": 7.285285285285286e-06,
+      "loss": 0.496,
+      "step": 27120
+    },
+    {
+      "epoch": 81.47,
+      "grad_norm": 15.112213134765625,
+      "learning_rate": 7.2842842842842855e-06,
+      "loss": 0.5332,
+      "step": 27130
+    },
+    {
+      "epoch": 81.5,
+      "grad_norm": 14.706315040588379,
+      "learning_rate": 7.283283283283283e-06,
+      "loss": 0.5166,
+      "step": 27140
+    },
+    {
+      "epoch": 81.53,
+      "grad_norm": 17.013132095336914,
+      "learning_rate": 7.282282282282283e-06,
+      "loss": 0.4587,
+      "step": 27150
+    },
+    {
+      "epoch": 81.56,
+      "grad_norm": 14.217673301696777,
+      "learning_rate": 7.281281281281281e-06,
+      "loss": 0.5012,
+      "step": 27160
+    },
+    {
+      "epoch": 81.59,
+      "grad_norm": 16.70499610900879,
+      "learning_rate": 7.280280280280281e-06,
+      "loss": 0.5238,
+      "step": 27170
+    },
+    {
+      "epoch": 81.62,
+      "grad_norm": 18.069900512695312,
+      "learning_rate": 7.27927927927928e-06,
+      "loss": 0.5143,
+      "step": 27180
+    },
+    {
+      "epoch": 81.65,
+      "grad_norm": 20.117631912231445,
+      "learning_rate": 7.278278278278279e-06,
+      "loss": 0.4739,
+      "step": 27190
+    },
+    {
+      "epoch": 81.68,
+      "grad_norm": 20.28511619567871,
+      "learning_rate": 7.277277277277278e-06,
+      "loss": 0.4929,
+      "step": 27200
+    },
+    {
+      "epoch": 81.71,
+      "grad_norm": 13.58857250213623,
+      "learning_rate": 7.276276276276278e-06,
+      "loss": 0.527,
+      "step": 27210
+    },
+    {
+      "epoch": 81.74,
+      "grad_norm": 12.338050842285156,
+      "learning_rate": 7.275275275275275e-06,
+      "loss": 0.4731,
+      "step": 27220
+    },
+    {
+      "epoch": 81.77,
+      "grad_norm": 10.727715492248535,
+      "learning_rate": 7.274274274274275e-06,
+      "loss": 0.6063,
+      "step": 27230
+    },
+    {
+      "epoch": 81.8,
+      "grad_norm": 19.270235061645508,
+      "learning_rate": 7.273273273273273e-06,
+      "loss": 0.5,
+      "step": 27240
+    },
+    {
+      "epoch": 81.83,
+      "grad_norm": 11.240653991699219,
+      "learning_rate": 7.272272272272273e-06,
+      "loss": 0.4847,
+      "step": 27250
+    },
+    {
+      "epoch": 81.86,
+      "grad_norm": 16.444652557373047,
+      "learning_rate": 7.271271271271272e-06,
+      "loss": 0.466,
+      "step": 27260
+    },
+    {
+      "epoch": 81.89,
+      "grad_norm": 12.542959213256836,
+      "learning_rate": 7.270270270270271e-06,
+      "loss": 0.5569,
+      "step": 27270
+    },
+    {
+      "epoch": 81.92,
+      "grad_norm": 16.11238670349121,
+      "learning_rate": 7.26926926926927e-06,
+      "loss": 0.5396,
+      "step": 27280
+    },
+    {
+      "epoch": 81.95,
+      "grad_norm": 23.949087142944336,
+      "learning_rate": 7.268268268268269e-06,
+      "loss": 0.5495,
+      "step": 27290
+    },
+    {
+      "epoch": 81.98,
+      "grad_norm": 17.89312744140625,
+      "learning_rate": 7.267267267267268e-06,
+      "loss": 0.4648,
+      "step": 27300
+    },
+    {
+      "epoch": 82.0,
+      "eval_accuracy": 0.8681,
+      "eval_loss": 0.4730146825313568,
+      "eval_runtime": 12.8199,
+      "eval_samples_per_second": 780.04,
+      "eval_steps_per_second": 3.12,
+      "step": 27306
+    },
+    {
+      "epoch": 82.01,
+      "grad_norm": 23.988683700561523,
+      "learning_rate": 7.266266266266266e-06,
+      "loss": 0.5477,
+      "step": 27310
+    },
+    {
+      "epoch": 82.04,
+      "grad_norm": 11.013201713562012,
+      "learning_rate": 7.2652652652652655e-06,
+      "loss": 0.4666,
+      "step": 27320
+    },
+    {
+      "epoch": 82.07,
+      "grad_norm": 14.66703987121582,
+      "learning_rate": 7.264264264264265e-06,
+      "loss": 0.467,
+      "step": 27330
+    },
+    {
+      "epoch": 82.1,
+      "grad_norm": 19.158966064453125,
+      "learning_rate": 7.263263263263264e-06,
+      "loss": 0.5321,
+      "step": 27340
+    },
+    {
+      "epoch": 82.13,
+      "grad_norm": 17.552034378051758,
+      "learning_rate": 7.262262262262263e-06,
+      "loss": 0.5071,
+      "step": 27350
+    },
+    {
+      "epoch": 82.16,
+      "grad_norm": 11.901301383972168,
+      "learning_rate": 7.2612612612612625e-06,
+      "loss": 0.4968,
+      "step": 27360
+    },
+    {
+      "epoch": 82.19,
+      "grad_norm": 14.851964950561523,
+      "learning_rate": 7.260260260260261e-06,
+      "loss": 0.5088,
+      "step": 27370
+    },
+    {
+      "epoch": 82.22,
+      "grad_norm": 13.52309799194336,
+      "learning_rate": 7.2592592592592605e-06,
+      "loss": 0.5089,
+      "step": 27380
+    },
+    {
+      "epoch": 82.25,
+      "grad_norm": 15.34189224243164,
+      "learning_rate": 7.258258258258258e-06,
+      "loss": 0.4828,
+      "step": 27390
+    },
+    {
+      "epoch": 82.28,
+      "grad_norm": 21.503339767456055,
+      "learning_rate": 7.257257257257258e-06,
+      "loss": 0.5525,
+      "step": 27400
+    },
+    {
+      "epoch": 82.31,
+      "grad_norm": 16.29681968688965,
+      "learning_rate": 7.256256256256256e-06,
+      "loss": 0.4926,
+      "step": 27410
+    },
+    {
+      "epoch": 82.34,
+      "grad_norm": 20.325592041015625,
+      "learning_rate": 7.255255255255256e-06,
+      "loss": 0.5453,
+      "step": 27420
+    },
+    {
+      "epoch": 82.37,
+      "grad_norm": 22.434194564819336,
+      "learning_rate": 7.254254254254255e-06,
+      "loss": 0.5411,
+      "step": 27430
+    },
+    {
+      "epoch": 82.4,
+      "grad_norm": 19.42127227783203,
+      "learning_rate": 7.253253253253254e-06,
+      "loss": 0.5606,
+      "step": 27440
+    },
+    {
+      "epoch": 82.43,
+      "grad_norm": 12.874723434448242,
+      "learning_rate": 7.252252252252253e-06,
+      "loss": 0.539,
+      "step": 27450
+    },
+    {
+      "epoch": 82.46,
+      "grad_norm": 13.633169174194336,
+      "learning_rate": 7.251251251251253e-06,
+      "loss": 0.4834,
+      "step": 27460
+    },
+    {
+      "epoch": 82.49,
+      "grad_norm": 16.47105598449707,
+      "learning_rate": 7.25025025025025e-06,
+      "loss": 0.4278,
+      "step": 27470
+    },
+    {
+      "epoch": 82.52,
+      "grad_norm": 14.252326965332031,
+      "learning_rate": 7.24924924924925e-06,
+      "loss": 0.4782,
+      "step": 27480
+    },
+    {
+      "epoch": 82.55,
+      "grad_norm": 15.690261840820312,
+      "learning_rate": 7.2482482482482484e-06,
+      "loss": 0.5556,
+      "step": 27490
+    },
+    {
+      "epoch": 82.58,
+      "grad_norm": 12.490242958068848,
+      "learning_rate": 7.247247247247248e-06,
+      "loss": 0.5361,
+      "step": 27500
+    },
+    {
+      "epoch": 82.61,
+      "grad_norm": 16.0650634765625,
+      "learning_rate": 7.2462462462462465e-06,
+      "loss": 0.5288,
+      "step": 27510
+    },
+    {
+      "epoch": 82.64,
+      "grad_norm": 16.423114776611328,
+      "learning_rate": 7.245245245245246e-06,
+      "loss": 0.5194,
+      "step": 27520
+    },
+    {
+      "epoch": 82.67,
+      "grad_norm": 14.915671348571777,
+      "learning_rate": 7.244244244244245e-06,
+      "loss": 0.5426,
+      "step": 27530
+    },
+    {
+      "epoch": 82.7,
+      "grad_norm": 19.004558563232422,
+      "learning_rate": 7.243243243243244e-06,
+      "loss": 0.5105,
+      "step": 27540
+    },
+    {
+      "epoch": 82.73,
+      "grad_norm": 13.798174858093262,
+      "learning_rate": 7.242242242242243e-06,
+      "loss": 0.5387,
+      "step": 27550
+    },
+    {
+      "epoch": 82.76,
+      "grad_norm": 19.4901180267334,
+      "learning_rate": 7.241241241241241e-06,
+      "loss": 0.4877,
+      "step": 27560
+    },
+    {
+      "epoch": 82.79,
+      "grad_norm": 15.883666038513184,
+      "learning_rate": 7.240240240240241e-06,
+      "loss": 0.5695,
+      "step": 27570
+    },
+    {
+      "epoch": 82.82,
+      "grad_norm": 14.284727096557617,
+      "learning_rate": 7.23923923923924e-06,
+      "loss": 0.5544,
+      "step": 27580
+    },
+    {
+      "epoch": 82.85,
+      "grad_norm": 12.819677352905273,
+      "learning_rate": 7.238238238238239e-06,
+      "loss": 0.5676,
+      "step": 27590
+    },
+    {
+      "epoch": 82.88,
+      "grad_norm": 18.853370666503906,
+      "learning_rate": 7.237237237237238e-06,
+      "loss": 0.4831,
+      "step": 27600
+    },
+    {
+      "epoch": 82.91,
+      "grad_norm": 20.63605308532715,
+      "learning_rate": 7.236236236236237e-06,
+      "loss": 0.5092,
+      "step": 27610
+    },
+    {
+      "epoch": 82.94,
+      "grad_norm": 15.875106811523438,
+      "learning_rate": 7.235235235235236e-06,
+      "loss": 0.4739,
+      "step": 27620
+    },
+    {
+      "epoch": 82.97,
+      "grad_norm": 20.798343658447266,
+      "learning_rate": 7.2342342342342355e-06,
+      "loss": 0.5853,
+      "step": 27630
+    },
+    {
+      "epoch": 83.0,
+      "eval_accuracy": 0.8656,
+      "eval_loss": 0.4780651926994324,
+      "eval_runtime": 12.8752,
+      "eval_samples_per_second": 776.688,
+      "eval_steps_per_second": 3.107,
+      "step": 27639
+    },
+    {
+      "epoch": 83.0,
+      "grad_norm": 11.860279083251953,
+      "learning_rate": 7.233233233233233e-06,
+      "loss": 0.505,
+      "step": 27640
+    },
+    {
+      "epoch": 83.03,
+      "grad_norm": 18.203495025634766,
+      "learning_rate": 7.232232232232233e-06,
+      "loss": 0.5636,
+      "step": 27650
+    },
+    {
+      "epoch": 83.06,
+      "grad_norm": 13.584324836730957,
+      "learning_rate": 7.231231231231231e-06,
+      "loss": 0.4661,
+      "step": 27660
+    },
+    {
+      "epoch": 83.09,
+      "grad_norm": 13.434273719787598,
+      "learning_rate": 7.230230230230231e-06,
+      "loss": 0.4884,
+      "step": 27670
+    },
+    {
+      "epoch": 83.12,
+      "grad_norm": 26.261144638061523,
+      "learning_rate": 7.22922922922923e-06,
+      "loss": 0.5101,
+      "step": 27680
+    },
+    {
+      "epoch": 83.15,
+      "grad_norm": 15.838640213012695,
+      "learning_rate": 7.228228228228229e-06,
+      "loss": 0.4783,
+      "step": 27690
+    },
+    {
+      "epoch": 83.18,
+      "grad_norm": 19.25486946105957,
+      "learning_rate": 7.227227227227228e-06,
+      "loss": 0.5055,
+      "step": 27700
+    },
+    {
+      "epoch": 83.21,
+      "grad_norm": 12.068188667297363,
+      "learning_rate": 7.226226226226228e-06,
+      "loss": 0.5237,
+      "step": 27710
+    },
+    {
+      "epoch": 83.24,
+      "grad_norm": 16.15995979309082,
+      "learning_rate": 7.2252252252252254e-06,
+      "loss": 0.4966,
+      "step": 27720
+    },
+    {
+      "epoch": 83.27,
+      "grad_norm": 18.45575714111328,
+      "learning_rate": 7.224224224224224e-06,
+      "loss": 0.5047,
+      "step": 27730
+    },
+    {
+      "epoch": 83.3,
+      "grad_norm": 21.30694580078125,
+      "learning_rate": 7.2232232232232235e-06,
+      "loss": 0.515,
+      "step": 27740
+    },
+    {
+      "epoch": 83.33,
+      "grad_norm": 14.493951797485352,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.508,
+      "step": 27750
+    },
+    {
+      "epoch": 83.36,
+      "grad_norm": 9.595751762390137,
+      "learning_rate": 7.2212212212212215e-06,
+      "loss": 0.4961,
+      "step": 27760
+    },
+    {
+      "epoch": 83.39,
+      "grad_norm": 19.376615524291992,
+      "learning_rate": 7.220220220220221e-06,
+      "loss": 0.5333,
+      "step": 27770
+    },
+    {
+      "epoch": 83.42,
+      "grad_norm": 19.98113250732422,
+      "learning_rate": 7.21921921921922e-06,
+      "loss": 0.4903,
+      "step": 27780
+    },
+    {
+      "epoch": 83.45,
+      "grad_norm": 12.427263259887695,
+      "learning_rate": 7.218218218218219e-06,
+      "loss": 0.4672,
+      "step": 27790
+    },
+    {
+      "epoch": 83.48,
+      "grad_norm": 16.322376251220703,
+      "learning_rate": 7.217217217217218e-06,
+      "loss": 0.5028,
+      "step": 27800
+    },
+    {
+      "epoch": 83.51,
+      "grad_norm": 16.018421173095703,
+      "learning_rate": 7.216216216216216e-06,
+      "loss": 0.5217,
+      "step": 27810
+    },
+    {
+      "epoch": 83.54,
+      "grad_norm": 10.323281288146973,
+      "learning_rate": 7.215215215215216e-06,
+      "loss": 0.4973,
+      "step": 27820
+    },
+    {
+      "epoch": 83.57,
+      "grad_norm": 10.037334442138672,
+      "learning_rate": 7.214214214214215e-06,
+      "loss": 0.5405,
+      "step": 27830
+    },
+    {
+      "epoch": 83.6,
+      "grad_norm": 13.62839412689209,
+      "learning_rate": 7.213213213213214e-06,
+      "loss": 0.509,
+      "step": 27840
+    },
+    {
+      "epoch": 83.63,
+      "grad_norm": 14.484914779663086,
+      "learning_rate": 7.212212212212213e-06,
+      "loss": 0.5128,
+      "step": 27850
+    },
+    {
+      "epoch": 83.66,
+      "grad_norm": 13.376684188842773,
+      "learning_rate": 7.211211211211212e-06,
+      "loss": 0.5263,
+      "step": 27860
+    },
+    {
+      "epoch": 83.69,
+      "grad_norm": 19.595674514770508,
+      "learning_rate": 7.210210210210211e-06,
+      "loss": 0.4579,
+      "step": 27870
+    },
+    {
+      "epoch": 83.72,
+      "grad_norm": 15.632843017578125,
+      "learning_rate": 7.2092092092092106e-06,
+      "loss": 0.4914,
+      "step": 27880
+    },
+    {
+      "epoch": 83.75,
+      "grad_norm": 16.465303421020508,
+      "learning_rate": 7.208208208208208e-06,
+      "loss": 0.5398,
+      "step": 27890
+    },
+    {
+      "epoch": 83.78,
+      "grad_norm": 19.066591262817383,
+      "learning_rate": 7.207207207207208e-06,
+      "loss": 0.4711,
+      "step": 27900
+    },
+    {
+      "epoch": 83.81,
+      "grad_norm": 15.890423774719238,
+      "learning_rate": 7.206206206206206e-06,
+      "loss": 0.56,
+      "step": 27910
+    },
+    {
+      "epoch": 83.84,
+      "grad_norm": 20.270509719848633,
+      "learning_rate": 7.205205205205206e-06,
+      "loss": 0.5004,
+      "step": 27920
+    },
+    {
+      "epoch": 83.87,
+      "grad_norm": 14.511930465698242,
+      "learning_rate": 7.204204204204205e-06,
+      "loss": 0.5083,
+      "step": 27930
+    },
+    {
+      "epoch": 83.9,
+      "grad_norm": 15.224581718444824,
+      "learning_rate": 7.203203203203204e-06,
+      "loss": 0.5081,
+      "step": 27940
+    },
+    {
+      "epoch": 83.93,
+      "grad_norm": 14.492236137390137,
+      "learning_rate": 7.202202202202203e-06,
+      "loss": 0.5216,
+      "step": 27950
+    },
+    {
+      "epoch": 83.96,
+      "grad_norm": 15.13267707824707,
+      "learning_rate": 7.201201201201202e-06,
+      "loss": 0.451,
+      "step": 27960
+    },
+    {
+      "epoch": 83.99,
+      "grad_norm": 20.824356079101562,
+      "learning_rate": 7.2002002002002005e-06,
+      "loss": 0.5298,
+      "step": 27970
+    },
+    {
+      "epoch": 84.0,
+      "eval_accuracy": 0.869,
+      "eval_loss": 0.4728562533855438,
+      "eval_runtime": 13.0251,
+      "eval_samples_per_second": 767.751,
+      "eval_steps_per_second": 3.071,
+      "step": 27972
+    },
+    {
+      "epoch": 84.02,
+      "grad_norm": 10.574320793151855,
+      "learning_rate": 7.199199199199199e-06,
+      "loss": 0.604,
+      "step": 27980
+    },
+    {
+      "epoch": 84.05,
+      "grad_norm": 14.731752395629883,
+      "learning_rate": 7.1981981981981985e-06,
+      "loss": 0.4887,
+      "step": 27990
+    },
+    {
+      "epoch": 84.08,
+      "grad_norm": 10.434557914733887,
+      "learning_rate": 7.197197197197198e-06,
+      "loss": 0.4773,
+      "step": 28000
+    },
+    {
+      "epoch": 84.11,
+      "grad_norm": 13.42420768737793,
+      "learning_rate": 7.1961961961961965e-06,
+      "loss": 0.5254,
+      "step": 28010
+    },
+    {
+      "epoch": 84.14,
+      "grad_norm": 13.98668098449707,
+      "learning_rate": 7.195195195195196e-06,
+      "loss": 0.4986,
+      "step": 28020
+    },
+    {
+      "epoch": 84.17,
+      "grad_norm": 12.513934135437012,
+      "learning_rate": 7.194194194194195e-06,
+      "loss": 0.4707,
+      "step": 28030
+    },
+    {
+      "epoch": 84.2,
+      "grad_norm": 15.887160301208496,
+      "learning_rate": 7.193193193193194e-06,
+      "loss": 0.4697,
+      "step": 28040
+    },
+    {
+      "epoch": 84.23,
+      "grad_norm": 15.136001586914062,
+      "learning_rate": 7.1921921921921935e-06,
+      "loss": 0.5131,
+      "step": 28050
+    },
+    {
+      "epoch": 84.26,
+      "grad_norm": 13.415595054626465,
+      "learning_rate": 7.191191191191191e-06,
+      "loss": 0.4972,
+      "step": 28060
+    },
+    {
+      "epoch": 84.29,
+      "grad_norm": 16.6083984375,
+      "learning_rate": 7.190190190190191e-06,
+      "loss": 0.4587,
+      "step": 28070
+    },
+    {
+      "epoch": 84.32,
+      "grad_norm": 28.22075843811035,
+      "learning_rate": 7.189189189189189e-06,
+      "loss": 0.5245,
+      "step": 28080
+    },
+    {
+      "epoch": 84.35,
+      "grad_norm": 12.82327938079834,
+      "learning_rate": 7.188188188188189e-06,
+      "loss": 0.5054,
+      "step": 28090
+    },
+    {
+      "epoch": 84.38,
+      "grad_norm": 17.963396072387695,
+      "learning_rate": 7.187187187187188e-06,
+      "loss": 0.5275,
+      "step": 28100
+    },
+    {
+      "epoch": 84.41,
+      "grad_norm": 13.487298965454102,
+      "learning_rate": 7.186186186186187e-06,
+      "loss": 0.4571,
+      "step": 28110
+    },
+    {
+      "epoch": 84.44,
+      "grad_norm": 17.037858963012695,
+      "learning_rate": 7.185185185185186e-06,
+      "loss": 0.5254,
+      "step": 28120
+    },
+    {
+      "epoch": 84.47,
+      "grad_norm": 16.970691680908203,
+      "learning_rate": 7.184184184184186e-06,
+      "loss": 0.4804,
+      "step": 28130
+    },
+    {
+      "epoch": 84.5,
+      "grad_norm": 19.333240509033203,
+      "learning_rate": 7.183183183183183e-06,
+      "loss": 0.5244,
+      "step": 28140
+    },
+    {
+      "epoch": 84.53,
+      "grad_norm": 12.293442726135254,
+      "learning_rate": 7.182182182182183e-06,
+      "loss": 0.4906,
+      "step": 28150
+    },
+    {
+      "epoch": 84.56,
+      "grad_norm": 13.594487190246582,
+      "learning_rate": 7.181181181181181e-06,
+      "loss": 0.4918,
+      "step": 28160
+    },
+    {
+      "epoch": 84.59,
+      "grad_norm": 13.262083053588867,
+      "learning_rate": 7.180180180180181e-06,
+      "loss": 0.5092,
+      "step": 28170
+    },
+    {
+      "epoch": 84.62,
+      "grad_norm": 20.580219268798828,
+      "learning_rate": 7.1791791791791794e-06,
+      "loss": 0.5616,
+      "step": 28180
+    },
+    {
+      "epoch": 84.65,
+      "grad_norm": 19.55314064025879,
+      "learning_rate": 7.178178178178179e-06,
+      "loss": 0.5074,
+      "step": 28190
+    },
+    {
+      "epoch": 84.68,
+      "grad_norm": 16.784488677978516,
+      "learning_rate": 7.177177177177178e-06,
+      "loss": 0.5366,
+      "step": 28200
+    },
+    {
+      "epoch": 84.71,
+      "grad_norm": 15.993558883666992,
+      "learning_rate": 7.176176176176177e-06,
+      "loss": 0.5448,
+      "step": 28210
+    },
+    {
+      "epoch": 84.74,
+      "grad_norm": 15.678027153015137,
+      "learning_rate": 7.1751751751751755e-06,
+      "loss": 0.5437,
+      "step": 28220
+    },
+    {
+      "epoch": 84.77,
+      "grad_norm": 11.770125389099121,
+      "learning_rate": 7.174174174174174e-06,
+      "loss": 0.535,
+      "step": 28230
+    },
+    {
+      "epoch": 84.8,
+      "grad_norm": 18.855430603027344,
+      "learning_rate": 7.1731731731731735e-06,
+      "loss": 0.5175,
+      "step": 28240
+    },
+    {
+      "epoch": 84.83,
+      "grad_norm": 23.09416389465332,
+      "learning_rate": 7.172172172172173e-06,
+      "loss": 0.5511,
+      "step": 28250
+    },
+    {
+      "epoch": 84.86,
+      "grad_norm": 15.866230964660645,
+      "learning_rate": 7.1711711711711716e-06,
+      "loss": 0.51,
+      "step": 28260
+    },
+    {
+      "epoch": 84.89,
+      "grad_norm": 16.555824279785156,
+      "learning_rate": 7.170170170170171e-06,
+      "loss": 0.4589,
+      "step": 28270
+    },
+    {
+      "epoch": 84.92,
+      "grad_norm": 13.851365089416504,
+      "learning_rate": 7.1691691691691705e-06,
+      "loss": 0.5211,
+      "step": 28280
+    },
+    {
+      "epoch": 84.95,
+      "grad_norm": 12.860268592834473,
+      "learning_rate": 7.168168168168169e-06,
+      "loss": 0.4454,
+      "step": 28290
+    },
+    {
+      "epoch": 84.98,
+      "grad_norm": 14.178278923034668,
+      "learning_rate": 7.167167167167167e-06,
+      "loss": 0.4484,
+      "step": 28300
+    },
+    {
+      "epoch": 85.0,
+      "eval_accuracy": 0.869,
+      "eval_loss": 0.4740825593471527,
+      "eval_runtime": 12.7553,
+      "eval_samples_per_second": 783.987,
+      "eval_steps_per_second": 3.136,
+      "step": 28305
+    },
+    {
+      "epoch": 85.02,
+      "grad_norm": 18.63173484802246,
+      "learning_rate": 7.166166166166166e-06,
+      "loss": 0.6098,
+      "step": 28310
+    },
+    {
+      "epoch": 85.05,
+      "grad_norm": 19.61794662475586,
+      "learning_rate": 7.165165165165166e-06,
+      "loss": 0.5282,
+      "step": 28320
+    },
+    {
+      "epoch": 85.08,
+      "grad_norm": 20.200681686401367,
+      "learning_rate": 7.164164164164164e-06,
+      "loss": 0.4699,
+      "step": 28330
+    },
+    {
+      "epoch": 85.11,
+      "grad_norm": 23.22762680053711,
+      "learning_rate": 7.163163163163164e-06,
+      "loss": 0.5258,
+      "step": 28340
+    },
+    {
+      "epoch": 85.14,
+      "grad_norm": 13.644779205322266,
+      "learning_rate": 7.162162162162163e-06,
+      "loss": 0.5212,
+      "step": 28350
+    },
+    {
+      "epoch": 85.17,
+      "grad_norm": 15.226165771484375,
+      "learning_rate": 7.161161161161162e-06,
+      "loss": 0.4739,
+      "step": 28360
+    },
+    {
+      "epoch": 85.2,
+      "grad_norm": 19.651981353759766,
+      "learning_rate": 7.160160160160161e-06,
+      "loss": 0.472,
+      "step": 28370
+    },
+    {
+      "epoch": 85.23,
+      "grad_norm": 12.74783992767334,
+      "learning_rate": 7.159159159159161e-06,
+      "loss": 0.4689,
+      "step": 28380
+    },
+    {
+      "epoch": 85.26,
+      "grad_norm": 13.573821067810059,
+      "learning_rate": 7.158158158158158e-06,
+      "loss": 0.5248,
+      "step": 28390
+    },
+    {
+      "epoch": 85.29,
+      "grad_norm": 17.988346099853516,
+      "learning_rate": 7.157157157157158e-06,
+      "loss": 0.5036,
+      "step": 28400
+    },
+    {
+      "epoch": 85.32,
+      "grad_norm": 16.88225746154785,
+      "learning_rate": 7.156156156156156e-06,
+      "loss": 0.5256,
+      "step": 28410
+    },
+    {
+      "epoch": 85.35,
+      "grad_norm": 17.99980926513672,
+      "learning_rate": 7.155155155155156e-06,
+      "loss": 0.5283,
+      "step": 28420
+    },
+    {
+      "epoch": 85.38,
+      "grad_norm": 16.545686721801758,
+      "learning_rate": 7.1541541541541545e-06,
+      "loss": 0.5181,
+      "step": 28430
+    },
+    {
+      "epoch": 85.41,
+      "grad_norm": 18.98824691772461,
+      "learning_rate": 7.153153153153154e-06,
+      "loss": 0.4657,
+      "step": 28440
+    },
+    {
+      "epoch": 85.44,
+      "grad_norm": 14.740886688232422,
+      "learning_rate": 7.152152152152153e-06,
+      "loss": 0.5669,
+      "step": 28450
+    },
+    {
+      "epoch": 85.47,
+      "grad_norm": 15.921751022338867,
+      "learning_rate": 7.151151151151152e-06,
+      "loss": 0.4818,
+      "step": 28460
+    },
+    {
+      "epoch": 85.5,
+      "grad_norm": 13.95246410369873,
+      "learning_rate": 7.1501501501501505e-06,
+      "loss": 0.4806,
+      "step": 28470
+    },
+    {
+      "epoch": 85.53,
+      "grad_norm": 14.615503311157227,
+      "learning_rate": 7.149149149149149e-06,
+      "loss": 0.4843,
+      "step": 28480
+    },
+    {
+      "epoch": 85.56,
+      "grad_norm": 22.90113639831543,
+      "learning_rate": 7.1481481481481486e-06,
+      "loss": 0.5608,
+      "step": 28490
+    },
+    {
+      "epoch": 85.59,
+      "grad_norm": 17.773834228515625,
+      "learning_rate": 7.147147147147148e-06,
+      "loss": 0.5189,
+      "step": 28500
+    },
+    {
+      "epoch": 85.62,
+      "grad_norm": 16.468957901000977,
+      "learning_rate": 7.146146146146147e-06,
+      "loss": 0.5018,
+      "step": 28510
+    },
+    {
+      "epoch": 85.65,
+      "grad_norm": 20.70461654663086,
+      "learning_rate": 7.145145145145146e-06,
+      "loss": 0.4561,
+      "step": 28520
+    },
+    {
+      "epoch": 85.68,
+      "grad_norm": 11.857874870300293,
+      "learning_rate": 7.144144144144145e-06,
+      "loss": 0.516,
+      "step": 28530
+    },
+    {
+      "epoch": 85.71,
+      "grad_norm": 13.240480422973633,
+      "learning_rate": 7.143143143143144e-06,
+      "loss": 0.4974,
+      "step": 28540
+    },
+    {
+      "epoch": 85.74,
+      "grad_norm": 14.336261749267578,
+      "learning_rate": 7.142142142142142e-06,
+      "loss": 0.4871,
+      "step": 28550
+    },
+    {
+      "epoch": 85.77,
+      "grad_norm": 9.17609977722168,
+      "learning_rate": 7.141141141141141e-06,
+      "loss": 0.4628,
+      "step": 28560
+    },
+    {
+      "epoch": 85.8,
+      "grad_norm": 12.698349952697754,
+      "learning_rate": 7.140140140140141e-06,
+      "loss": 0.5396,
+      "step": 28570
+    },
+    {
+      "epoch": 85.83,
+      "grad_norm": 17.80215072631836,
+      "learning_rate": 7.139139139139139e-06,
+      "loss": 0.4676,
+      "step": 28580
+    },
+    {
+      "epoch": 85.86,
+      "grad_norm": 20.226518630981445,
+      "learning_rate": 7.138138138138139e-06,
+      "loss": 0.5154,
+      "step": 28590
+    },
+    {
+      "epoch": 85.89,
+      "grad_norm": 15.085134506225586,
+      "learning_rate": 7.137137137137138e-06,
+      "loss": 0.494,
+      "step": 28600
+    },
+    {
+      "epoch": 85.92,
+      "grad_norm": 12.757584571838379,
+      "learning_rate": 7.136136136136137e-06,
+      "loss": 0.537,
+      "step": 28610
+    },
+    {
+      "epoch": 85.95,
+      "grad_norm": 12.16426944732666,
+      "learning_rate": 7.135135135135136e-06,
+      "loss": 0.4782,
+      "step": 28620
+    },
+    {
+      "epoch": 85.98,
+      "grad_norm": 12.735384941101074,
+      "learning_rate": 7.134134134134135e-06,
+      "loss": 0.4765,
+      "step": 28630
+    },
+    {
+      "epoch": 86.0,
+      "eval_accuracy": 0.8633,
+      "eval_loss": 0.4877474009990692,
+      "eval_runtime": 12.5121,
+      "eval_samples_per_second": 799.225,
+      "eval_steps_per_second": 3.197,
+      "step": 28638
+    },
+    {
+      "epoch": 86.01,
+      "grad_norm": 20.190876007080078,
+      "learning_rate": 7.133133133133133e-06,
+      "loss": 0.5358,
+      "step": 28640
+    },
+    {
+      "epoch": 86.04,
+      "grad_norm": 16.767045974731445,
+      "learning_rate": 7.132132132132132e-06,
+      "loss": 0.5106,
+      "step": 28650
+    },
+    {
+      "epoch": 86.07,
+      "grad_norm": 16.47939682006836,
+      "learning_rate": 7.1311311311311315e-06,
+      "loss": 0.5083,
+      "step": 28660
+    },
+    {
+      "epoch": 86.1,
+      "grad_norm": 12.725757598876953,
+      "learning_rate": 7.130130130130131e-06,
+      "loss": 0.4462,
+      "step": 28670
+    },
+    {
+      "epoch": 86.13,
+      "grad_norm": 21.844676971435547,
+      "learning_rate": 7.1291291291291295e-06,
+      "loss": 0.4942,
+      "step": 28680
+    },
+    {
+      "epoch": 86.16,
+      "grad_norm": 13.987099647521973,
+      "learning_rate": 7.128128128128129e-06,
+      "loss": 0.4893,
+      "step": 28690
+    },
+    {
+      "epoch": 86.19,
+      "grad_norm": 12.143754959106445,
+      "learning_rate": 7.127127127127128e-06,
+      "loss": 0.4771,
+      "step": 28700
+    },
+    {
+      "epoch": 86.22,
+      "grad_norm": 19.128433227539062,
+      "learning_rate": 7.126126126126127e-06,
+      "loss": 0.503,
+      "step": 28710
+    },
+    {
+      "epoch": 86.25,
+      "grad_norm": 16.6035099029541,
+      "learning_rate": 7.1251251251251256e-06,
+      "loss": 0.5514,
+      "step": 28720
+    },
+    {
+      "epoch": 86.28,
+      "grad_norm": 16.854846954345703,
+      "learning_rate": 7.124124124124124e-06,
+      "loss": 0.4651,
+      "step": 28730
+    },
+    {
+      "epoch": 86.31,
+      "grad_norm": 13.618094444274902,
+      "learning_rate": 7.123123123123124e-06,
+      "loss": 0.5122,
+      "step": 28740
+    },
+    {
+      "epoch": 86.34,
+      "grad_norm": 14.200691223144531,
+      "learning_rate": 7.122122122122122e-06,
+      "loss": 0.5339,
+      "step": 28750
+    },
+    {
+      "epoch": 86.37,
+      "grad_norm": 14.8397855758667,
+      "learning_rate": 7.121121121121122e-06,
+      "loss": 0.5011,
+      "step": 28760
+    },
+    {
+      "epoch": 86.4,
+      "grad_norm": 11.214278221130371,
+      "learning_rate": 7.120120120120121e-06,
+      "loss": 0.5291,
+      "step": 28770
+    },
+    {
+      "epoch": 86.43,
+      "grad_norm": 18.572744369506836,
+      "learning_rate": 7.11911911911912e-06,
+      "loss": 0.5285,
+      "step": 28780
+    },
+    {
+      "epoch": 86.46,
+      "grad_norm": 18.296024322509766,
+      "learning_rate": 7.118118118118119e-06,
+      "loss": 0.5268,
+      "step": 28790
+    },
+    {
+      "epoch": 86.49,
+      "grad_norm": 13.891396522521973,
+      "learning_rate": 7.117117117117117e-06,
+      "loss": 0.5011,
+      "step": 28800
+    },
+    {
+      "epoch": 86.52,
+      "grad_norm": 14.529227256774902,
+      "learning_rate": 7.116116116116116e-06,
+      "loss": 0.4931,
+      "step": 28810
+    },
+    {
+      "epoch": 86.55,
+      "grad_norm": 17.182336807250977,
+      "learning_rate": 7.115115115115116e-06,
+      "loss": 0.4914,
+      "step": 28820
+    },
+    {
+      "epoch": 86.58,
+      "grad_norm": 17.16726303100586,
+      "learning_rate": 7.114114114114114e-06,
+      "loss": 0.5218,
+      "step": 28830
+    },
+    {
+      "epoch": 86.61,
+      "grad_norm": 14.171222686767578,
+      "learning_rate": 7.113113113113114e-06,
+      "loss": 0.5402,
+      "step": 28840
+    },
+    {
+      "epoch": 86.64,
+      "grad_norm": 21.08846664428711,
+      "learning_rate": 7.112112112112113e-06,
+      "loss": 0.5499,
+      "step": 28850
+    },
+    {
+      "epoch": 86.67,
+      "grad_norm": 17.825288772583008,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 0.4919,
+      "step": 28860
+    },
+    {
+      "epoch": 86.7,
+      "grad_norm": 15.799781799316406,
+      "learning_rate": 7.110110110110111e-06,
+      "loss": 0.5057,
+      "step": 28870
+    },
+    {
+      "epoch": 86.73,
+      "grad_norm": 10.760543823242188,
+      "learning_rate": 7.10910910910911e-06,
+      "loss": 0.4696,
+      "step": 28880
+    },
+    {
+      "epoch": 86.76,
+      "grad_norm": 12.362149238586426,
+      "learning_rate": 7.1081081081081085e-06,
+      "loss": 0.5351,
+      "step": 28890
+    },
+    {
+      "epoch": 86.79,
+      "grad_norm": 22.476179122924805,
+      "learning_rate": 7.107107107107107e-06,
+      "loss": 0.5215,
+      "step": 28900
+    },
+    {
+      "epoch": 86.82,
+      "grad_norm": 12.070683479309082,
+      "learning_rate": 7.1061061061061065e-06,
+      "loss": 0.4834,
+      "step": 28910
+    },
+    {
+      "epoch": 86.85,
+      "grad_norm": 18.42873191833496,
+      "learning_rate": 7.105105105105106e-06,
+      "loss": 0.4361,
+      "step": 28920
+    },
+    {
+      "epoch": 86.88,
+      "grad_norm": 25.31146812438965,
+      "learning_rate": 7.1041041041041045e-06,
+      "loss": 0.5506,
+      "step": 28930
+    },
+    {
+      "epoch": 86.91,
+      "grad_norm": 13.773992538452148,
+      "learning_rate": 7.103103103103104e-06,
+      "loss": 0.5195,
+      "step": 28940
+    },
+    {
+      "epoch": 86.94,
+      "grad_norm": 20.223474502563477,
+      "learning_rate": 7.102102102102103e-06,
+      "loss": 0.4811,
+      "step": 28950
+    },
+    {
+      "epoch": 86.97,
+      "grad_norm": 13.442011833190918,
+      "learning_rate": 7.101101101101102e-06,
+      "loss": 0.4472,
+      "step": 28960
+    },
+    {
+      "epoch": 87.0,
+      "grad_norm": 18.693674087524414,
+      "learning_rate": 7.100100100100101e-06,
+      "loss": 0.5409,
+      "step": 28970
+    },
+    {
+      "epoch": 87.0,
+      "eval_accuracy": 0.8664,
+      "eval_loss": 0.4806751012802124,
+      "eval_runtime": 12.7527,
+      "eval_samples_per_second": 784.148,
+      "eval_steps_per_second": 3.137,
+      "step": 28971
+    },
+    {
+      "epoch": 87.03,
+      "grad_norm": 11.919361114501953,
+      "learning_rate": 7.099099099099099e-06,
+      "loss": 0.445,
+      "step": 28980
+    },
+    {
+      "epoch": 87.06,
+      "grad_norm": 19.516067504882812,
+      "learning_rate": 7.098098098098099e-06,
+      "loss": 0.5561,
+      "step": 28990
+    },
+    {
+      "epoch": 87.09,
+      "grad_norm": 13.301137924194336,
+      "learning_rate": 7.097097097097097e-06,
+      "loss": 0.443,
+      "step": 29000
+    },
+    {
+      "epoch": 87.12,
+      "grad_norm": 17.93448829650879,
+      "learning_rate": 7.096096096096097e-06,
+      "loss": 0.4778,
+      "step": 29010
+    },
+    {
+      "epoch": 87.15,
+      "grad_norm": 13.911527633666992,
+      "learning_rate": 7.095095095095096e-06,
+      "loss": 0.4956,
+      "step": 29020
+    },
+    {
+      "epoch": 87.18,
+      "grad_norm": 19.80583953857422,
+      "learning_rate": 7.094094094094095e-06,
+      "loss": 0.4452,
+      "step": 29030
+    },
+    {
+      "epoch": 87.21,
+      "grad_norm": 15.735163688659668,
+      "learning_rate": 7.093093093093094e-06,
+      "loss": 0.5091,
+      "step": 29040
+    },
+    {
+      "epoch": 87.24,
+      "grad_norm": 28.821273803710938,
+      "learning_rate": 7.092092092092092e-06,
+      "loss": 0.5307,
+      "step": 29050
+    },
+    {
+      "epoch": 87.27,
+      "grad_norm": 30.416255950927734,
+      "learning_rate": 7.091091091091091e-06,
+      "loss": 0.5255,
+      "step": 29060
+    },
+    {
+      "epoch": 87.3,
+      "grad_norm": 15.77895450592041,
+      "learning_rate": 7.090090090090091e-06,
+      "loss": 0.4988,
+      "step": 29070
+    },
+    {
+      "epoch": 87.33,
+      "grad_norm": 17.492530822753906,
+      "learning_rate": 7.089089089089089e-06,
+      "loss": 0.4559,
+      "step": 29080
+    },
+    {
+      "epoch": 87.36,
+      "grad_norm": 18.281818389892578,
+      "learning_rate": 7.088088088088089e-06,
+      "loss": 0.4955,
+      "step": 29090
+    },
+    {
+      "epoch": 87.39,
+      "grad_norm": 17.973648071289062,
+      "learning_rate": 7.087087087087087e-06,
+      "loss": 0.4519,
+      "step": 29100
+    },
+    {
+      "epoch": 87.42,
+      "grad_norm": 13.263154029846191,
+      "learning_rate": 7.086086086086087e-06,
+      "loss": 0.4735,
+      "step": 29110
+    },
+    {
+      "epoch": 87.45,
+      "grad_norm": 14.076695442199707,
+      "learning_rate": 7.085085085085086e-06,
+      "loss": 0.5984,
+      "step": 29120
+    },
+    {
+      "epoch": 87.48,
+      "grad_norm": 11.15629768371582,
+      "learning_rate": 7.084084084084085e-06,
+      "loss": 0.5091,
+      "step": 29130
+    },
+    {
+      "epoch": 87.51,
+      "grad_norm": 13.458136558532715,
+      "learning_rate": 7.0830830830830835e-06,
+      "loss": 0.5102,
+      "step": 29140
+    },
+    {
+      "epoch": 87.54,
+      "grad_norm": 15.646551132202148,
+      "learning_rate": 7.082082082082082e-06,
+      "loss": 0.5237,
+      "step": 29150
+    },
+    {
+      "epoch": 87.57,
+      "grad_norm": 16.076602935791016,
+      "learning_rate": 7.0810810810810815e-06,
+      "loss": 0.4742,
+      "step": 29160
+    },
+    {
+      "epoch": 87.6,
+      "grad_norm": 17.216720581054688,
+      "learning_rate": 7.080080080080081e-06,
+      "loss": 0.4867,
+      "step": 29170
+    },
+    {
+      "epoch": 87.63,
+      "grad_norm": 15.960564613342285,
+      "learning_rate": 7.0790790790790796e-06,
+      "loss": 0.526,
+      "step": 29180
+    },
+    {
+      "epoch": 87.66,
+      "grad_norm": 12.131356239318848,
+      "learning_rate": 7.078078078078079e-06,
+      "loss": 0.5324,
+      "step": 29190
+    },
+    {
+      "epoch": 87.69,
+      "grad_norm": 14.858342170715332,
+      "learning_rate": 7.0770770770770784e-06,
+      "loss": 0.4377,
+      "step": 29200
+    },
+    {
+      "epoch": 87.72,
+      "grad_norm": 18.87102508544922,
+      "learning_rate": 7.076076076076077e-06,
+      "loss": 0.5009,
+      "step": 29210
+    },
+    {
+      "epoch": 87.75,
+      "grad_norm": 17.45355987548828,
+      "learning_rate": 7.075075075075075e-06,
+      "loss": 0.5376,
+      "step": 29220
+    },
+    {
+      "epoch": 87.78,
+      "grad_norm": 14.351983070373535,
+      "learning_rate": 7.074074074074074e-06,
+      "loss": 0.5263,
+      "step": 29230
+    },
+    {
+      "epoch": 87.81,
+      "grad_norm": 11.981090545654297,
+      "learning_rate": 7.073073073073074e-06,
+      "loss": 0.4976,
+      "step": 29240
+    },
+    {
+      "epoch": 87.84,
+      "grad_norm": 15.420083999633789,
+      "learning_rate": 7.072072072072072e-06,
+      "loss": 0.5039,
+      "step": 29250
+    },
+    {
+      "epoch": 87.87,
+      "grad_norm": 13.468639373779297,
+      "learning_rate": 7.071071071071072e-06,
+      "loss": 0.4937,
+      "step": 29260
+    },
+    {
+      "epoch": 87.9,
+      "grad_norm": 17.320049285888672,
+      "learning_rate": 7.070070070070071e-06,
+      "loss": 0.4758,
+      "step": 29270
+    },
+    {
+      "epoch": 87.93,
+      "grad_norm": 15.816226959228516,
+      "learning_rate": 7.06906906906907e-06,
+      "loss": 0.4588,
+      "step": 29280
+    },
+    {
+      "epoch": 87.96,
+      "grad_norm": 25.267274856567383,
+      "learning_rate": 7.068068068068069e-06,
+      "loss": 0.5345,
+      "step": 29290
+    },
+    {
+      "epoch": 87.99,
+      "grad_norm": 17.635854721069336,
+      "learning_rate": 7.067067067067067e-06,
+      "loss": 0.4778,
+      "step": 29300
+    },
+    {
+      "epoch": 88.0,
+      "eval_accuracy": 0.8677,
+      "eval_loss": 0.475267231464386,
+      "eval_runtime": 12.4175,
+      "eval_samples_per_second": 805.314,
+      "eval_steps_per_second": 3.221,
+      "step": 29304
+    },
+    {
+      "epoch": 88.02,
+      "grad_norm": 19.624452590942383,
+      "learning_rate": 7.066066066066066e-06,
+      "loss": 0.5788,
+      "step": 29310
+    },
+    {
+      "epoch": 88.05,
+      "grad_norm": 13.224385261535645,
+      "learning_rate": 7.065065065065066e-06,
+      "loss": 0.5006,
+      "step": 29320
+    },
+    {
+      "epoch": 88.08,
+      "grad_norm": 11.519399642944336,
+      "learning_rate": 7.064064064064064e-06,
+      "loss": 0.4921,
+      "step": 29330
+    },
+    {
+      "epoch": 88.11,
+      "grad_norm": 18.72711944580078,
+      "learning_rate": 7.063063063063064e-06,
+      "loss": 0.512,
+      "step": 29340
+    },
+    {
+      "epoch": 88.14,
+      "grad_norm": 18.440258026123047,
+      "learning_rate": 7.0620620620620625e-06,
+      "loss": 0.4926,
+      "step": 29350
+    },
+    {
+      "epoch": 88.17,
+      "grad_norm": 15.627878189086914,
+      "learning_rate": 7.061061061061062e-06,
+      "loss": 0.4521,
+      "step": 29360
+    },
+    {
+      "epoch": 88.2,
+      "grad_norm": 17.182098388671875,
+      "learning_rate": 7.060060060060061e-06,
+      "loss": 0.4928,
+      "step": 29370
+    },
+    {
+      "epoch": 88.23,
+      "grad_norm": 19.698566436767578,
+      "learning_rate": 7.059059059059059e-06,
+      "loss": 0.5339,
+      "step": 29380
+    },
+    {
+      "epoch": 88.26,
+      "grad_norm": 17.014854431152344,
+      "learning_rate": 7.0580580580580585e-06,
+      "loss": 0.488,
+      "step": 29390
+    },
+    {
+      "epoch": 88.29,
+      "grad_norm": 15.220260620117188,
+      "learning_rate": 7.057057057057057e-06,
+      "loss": 0.4544,
+      "step": 29400
+    },
+    {
+      "epoch": 88.32,
+      "grad_norm": 14.810593605041504,
+      "learning_rate": 7.0560560560560566e-06,
+      "loss": 0.4522,
+      "step": 29410
+    },
+    {
+      "epoch": 88.35,
+      "grad_norm": 18.512983322143555,
+      "learning_rate": 7.055055055055056e-06,
+      "loss": 0.5206,
+      "step": 29420
+    },
+    {
+      "epoch": 88.38,
+      "grad_norm": 20.590789794921875,
+      "learning_rate": 7.054054054054055e-06,
+      "loss": 0.4482,
+      "step": 29430
+    },
+    {
+      "epoch": 88.41,
+      "grad_norm": 12.691764831542969,
+      "learning_rate": 7.053053053053054e-06,
+      "loss": 0.459,
+      "step": 29440
+    },
+    {
+      "epoch": 88.44,
+      "grad_norm": 14.093256950378418,
+      "learning_rate": 7.052052052052053e-06,
+      "loss": 0.4907,
+      "step": 29450
+    },
+    {
+      "epoch": 88.47,
+      "grad_norm": 15.214815139770508,
+      "learning_rate": 7.051051051051052e-06,
+      "loss": 0.5184,
+      "step": 29460
+    },
+    {
+      "epoch": 88.5,
+      "grad_norm": 12.073654174804688,
+      "learning_rate": 7.05005005005005e-06,
+      "loss": 0.5084,
+      "step": 29470
+    },
+    {
+      "epoch": 88.53,
+      "grad_norm": 15.3258695602417,
+      "learning_rate": 7.049049049049049e-06,
+      "loss": 0.5069,
+      "step": 29480
+    },
+    {
+      "epoch": 88.56,
+      "grad_norm": 16.83928680419922,
+      "learning_rate": 7.048048048048049e-06,
+      "loss": 0.5081,
+      "step": 29490
+    },
+    {
+      "epoch": 88.59,
+      "grad_norm": 15.47718334197998,
+      "learning_rate": 7.047047047047047e-06,
+      "loss": 0.4896,
+      "step": 29500
+    },
+    {
+      "epoch": 88.62,
+      "grad_norm": 11.8483247756958,
+      "learning_rate": 7.046046046046047e-06,
+      "loss": 0.5388,
+      "step": 29510
+    },
+    {
+      "epoch": 88.65,
+      "grad_norm": 16.320606231689453,
+      "learning_rate": 7.045045045045046e-06,
+      "loss": 0.462,
+      "step": 29520
+    },
+    {
+      "epoch": 88.68,
+      "grad_norm": 15.262286186218262,
+      "learning_rate": 7.044044044044045e-06,
+      "loss": 0.4784,
+      "step": 29530
+    },
+    {
+      "epoch": 88.71,
+      "grad_norm": 12.162701606750488,
+      "learning_rate": 7.043043043043044e-06,
+      "loss": 0.5195,
+      "step": 29540
+    },
+    {
+      "epoch": 88.74,
+      "grad_norm": 15.841780662536621,
+      "learning_rate": 7.042042042042042e-06,
+      "loss": 0.5207,
+      "step": 29550
+    },
+    {
+      "epoch": 88.77,
+      "grad_norm": 13.804239273071289,
+      "learning_rate": 7.041041041041041e-06,
+      "loss": 0.5029,
+      "step": 29560
+    },
+    {
+      "epoch": 88.8,
+      "grad_norm": 12.5025634765625,
+      "learning_rate": 7.04004004004004e-06,
+      "loss": 0.4648,
+      "step": 29570
+    },
+    {
+      "epoch": 88.83,
+      "grad_norm": 15.818353652954102,
+      "learning_rate": 7.0390390390390395e-06,
+      "loss": 0.4804,
+      "step": 29580
+    },
+    {
+      "epoch": 88.86,
+      "grad_norm": 13.170658111572266,
+      "learning_rate": 7.038038038038039e-06,
+      "loss": 0.493,
+      "step": 29590
+    },
+    {
+      "epoch": 88.89,
+      "grad_norm": 12.559015274047852,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.5041,
+      "step": 29600
+    },
+    {
+      "epoch": 88.92,
+      "grad_norm": 10.675616264343262,
+      "learning_rate": 7.036036036036037e-06,
+      "loss": 0.5035,
+      "step": 29610
+    },
+    {
+      "epoch": 88.95,
+      "grad_norm": 13.700398445129395,
+      "learning_rate": 7.035035035035036e-06,
+      "loss": 0.4693,
+      "step": 29620
+    },
+    {
+      "epoch": 88.98,
+      "grad_norm": 20.062053680419922,
+      "learning_rate": 7.034034034034034e-06,
+      "loss": 0.508,
+      "step": 29630
+    },
+    {
+      "epoch": 89.0,
+      "eval_accuracy": 0.867,
+      "eval_loss": 0.4750248193740845,
+      "eval_runtime": 12.8132,
+      "eval_samples_per_second": 780.446,
+      "eval_steps_per_second": 3.122,
+      "step": 29637
+    },
+    {
+      "epoch": 89.01,
+      "grad_norm": 16.882631301879883,
+      "learning_rate": 7.0330330330330336e-06,
+      "loss": 0.3747,
+      "step": 29640
+    },
+    {
+      "epoch": 89.04,
+      "grad_norm": 16.249284744262695,
+      "learning_rate": 7.032032032032032e-06,
+      "loss": 0.5164,
+      "step": 29650
+    },
+    {
+      "epoch": 89.07,
+      "grad_norm": 26.293668746948242,
+      "learning_rate": 7.031031031031032e-06,
+      "loss": 0.4515,
+      "step": 29660
+    },
+    {
+      "epoch": 89.1,
+      "grad_norm": 22.30623435974121,
+      "learning_rate": 7.03003003003003e-06,
+      "loss": 0.5377,
+      "step": 29670
+    },
+    {
+      "epoch": 89.13,
+      "grad_norm": 19.941524505615234,
+      "learning_rate": 7.02902902902903e-06,
+      "loss": 0.5314,
+      "step": 29680
+    },
+    {
+      "epoch": 89.16,
+      "grad_norm": 15.501513481140137,
+      "learning_rate": 7.028028028028029e-06,
+      "loss": 0.445,
+      "step": 29690
+    },
+    {
+      "epoch": 89.19,
+      "grad_norm": 19.918249130249023,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.5079,
+      "step": 29700
+    },
+    {
+      "epoch": 89.22,
+      "grad_norm": 13.900308609008789,
+      "learning_rate": 7.026026026026027e-06,
+      "loss": 0.4472,
+      "step": 29710
+    },
+    {
+      "epoch": 89.25,
+      "grad_norm": 11.860645294189453,
+      "learning_rate": 7.025025025025025e-06,
+      "loss": 0.4591,
+      "step": 29720
+    },
+    {
+      "epoch": 89.28,
+      "grad_norm": 15.179025650024414,
+      "learning_rate": 7.024024024024024e-06,
+      "loss": 0.5286,
+      "step": 29730
+    },
+    {
+      "epoch": 89.31,
+      "grad_norm": 17.204662322998047,
+      "learning_rate": 7.023023023023024e-06,
+      "loss": 0.4999,
+      "step": 29740
+    },
+    {
+      "epoch": 89.34,
+      "grad_norm": 12.837636947631836,
+      "learning_rate": 7.022022022022022e-06,
+      "loss": 0.4789,
+      "step": 29750
+    },
+    {
+      "epoch": 89.37,
+      "grad_norm": 21.773744583129883,
+      "learning_rate": 7.021021021021022e-06,
+      "loss": 0.438,
+      "step": 29760
+    },
+    {
+      "epoch": 89.4,
+      "grad_norm": 25.687503814697266,
+      "learning_rate": 7.020020020020021e-06,
+      "loss": 0.5505,
+      "step": 29770
+    },
+    {
+      "epoch": 89.43,
+      "grad_norm": 19.49062728881836,
+      "learning_rate": 7.01901901901902e-06,
+      "loss": 0.4961,
+      "step": 29780
+    },
+    {
+      "epoch": 89.46,
+      "grad_norm": 13.12169075012207,
+      "learning_rate": 7.018018018018019e-06,
+      "loss": 0.494,
+      "step": 29790
+    },
+    {
+      "epoch": 89.49,
+      "grad_norm": 17.911029815673828,
+      "learning_rate": 7.017017017017017e-06,
+      "loss": 0.502,
+      "step": 29800
+    },
+    {
+      "epoch": 89.52,
+      "grad_norm": 43.872135162353516,
+      "learning_rate": 7.0160160160160164e-06,
+      "loss": 0.5371,
+      "step": 29810
+    },
+    {
+      "epoch": 89.55,
+      "grad_norm": 9.673543930053711,
+      "learning_rate": 7.015015015015015e-06,
+      "loss": 0.4873,
+      "step": 29820
+    },
+    {
+      "epoch": 89.58,
+      "grad_norm": 14.422541618347168,
+      "learning_rate": 7.0140140140140145e-06,
+      "loss": 0.4758,
+      "step": 29830
+    },
+    {
+      "epoch": 89.61,
+      "grad_norm": 11.973191261291504,
+      "learning_rate": 7.013013013013014e-06,
+      "loss": 0.4872,
+      "step": 29840
+    },
+    {
+      "epoch": 89.64,
+      "grad_norm": 13.877923965454102,
+      "learning_rate": 7.0120120120120125e-06,
+      "loss": 0.4936,
+      "step": 29850
+    },
+    {
+      "epoch": 89.67,
+      "grad_norm": 15.526756286621094,
+      "learning_rate": 7.011011011011012e-06,
+      "loss": 0.49,
+      "step": 29860
+    },
+    {
+      "epoch": 89.7,
+      "grad_norm": 33.905311584472656,
+      "learning_rate": 7.010010010010011e-06,
+      "loss": 0.4935,
+      "step": 29870
+    },
+    {
+      "epoch": 89.73,
+      "grad_norm": 16.815303802490234,
+      "learning_rate": 7.009009009009009e-06,
+      "loss": 0.5419,
+      "step": 29880
+    },
+    {
+      "epoch": 89.76,
+      "grad_norm": 16.048723220825195,
+      "learning_rate": 7.008008008008009e-06,
+      "loss": 0.444,
+      "step": 29890
+    },
+    {
+      "epoch": 89.79,
+      "grad_norm": 27.505531311035156,
+      "learning_rate": 7.007007007007007e-06,
+      "loss": 0.514,
+      "step": 29900
+    },
+    {
+      "epoch": 89.82,
+      "grad_norm": 14.29491138458252,
+      "learning_rate": 7.006006006006007e-06,
+      "loss": 0.4577,
+      "step": 29910
+    },
+    {
+      "epoch": 89.85,
+      "grad_norm": 13.674175262451172,
+      "learning_rate": 7.005005005005005e-06,
+      "loss": 0.498,
+      "step": 29920
+    },
+    {
+      "epoch": 89.88,
+      "grad_norm": 16.51004981994629,
+      "learning_rate": 7.004004004004005e-06,
+      "loss": 0.5176,
+      "step": 29930
+    },
+    {
+      "epoch": 89.91,
+      "grad_norm": 10.969858169555664,
+      "learning_rate": 7.003003003003004e-06,
+      "loss": 0.4456,
+      "step": 29940
+    },
+    {
+      "epoch": 89.94,
+      "grad_norm": 15.421603202819824,
+      "learning_rate": 7.002002002002003e-06,
+      "loss": 0.4372,
+      "step": 29950
+    },
+    {
+      "epoch": 89.97,
+      "grad_norm": 12.754996299743652,
+      "learning_rate": 7.001001001001002e-06,
+      "loss": 0.4877,
+      "step": 29960
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 64.5278549194336,
+      "learning_rate": 7e-06,
+      "loss": 0.4567,
+      "step": 29970
+    },
+    {
+      "epoch": 90.0,
+      "eval_accuracy": 0.8681,
+      "eval_loss": 0.48157060146331787,
+      "eval_runtime": 12.9235,
+      "eval_samples_per_second": 773.787,
+      "eval_steps_per_second": 3.095,
+      "step": 29970
+    },
+    {
+      "epoch": 90.03,
+      "grad_norm": 9.665785789489746,
+      "learning_rate": 6.998998998998999e-06,
+      "loss": 0.446,
+      "step": 29980
+    },
+    {
+      "epoch": 90.06,
+      "grad_norm": 31.673931121826172,
+      "learning_rate": 6.997997997997999e-06,
+      "loss": 0.496,
+      "step": 29990
+    },
+    {
+      "epoch": 90.09,
+      "grad_norm": 16.958635330200195,
+      "learning_rate": 6.996996996996997e-06,
+      "loss": 0.5747,
+      "step": 30000
+    },
+    {
+      "epoch": 90.12,
+      "grad_norm": 16.371265411376953,
+      "learning_rate": 6.995995995995997e-06,
+      "loss": 0.5298,
+      "step": 30010
+    },
+    {
+      "epoch": 90.15,
+      "grad_norm": 15.15768814086914,
+      "learning_rate": 6.994994994994995e-06,
+      "loss": 0.4442,
+      "step": 30020
+    },
+    {
+      "epoch": 90.18,
+      "grad_norm": 16.892885208129883,
+      "learning_rate": 6.993993993993995e-06,
+      "loss": 0.4983,
+      "step": 30030
+    },
+    {
+      "epoch": 90.21,
+      "grad_norm": 15.908172607421875,
+      "learning_rate": 6.992992992992994e-06,
+      "loss": 0.5088,
+      "step": 30040
+    },
+    {
+      "epoch": 90.24,
+      "grad_norm": 23.386457443237305,
+      "learning_rate": 6.991991991991992e-06,
+      "loss": 0.4715,
+      "step": 30050
+    },
+    {
+      "epoch": 90.27,
+      "grad_norm": 14.932313919067383,
+      "learning_rate": 6.9909909909909915e-06,
+      "loss": 0.499,
+      "step": 30060
+    },
+    {
+      "epoch": 90.3,
+      "grad_norm": 16.232872009277344,
+      "learning_rate": 6.98998998998999e-06,
+      "loss": 0.514,
+      "step": 30070
+    },
+    {
+      "epoch": 90.33,
+      "grad_norm": 10.346344947814941,
+      "learning_rate": 6.9889889889889895e-06,
+      "loss": 0.5141,
+      "step": 30080
+    },
+    {
+      "epoch": 90.36,
+      "grad_norm": 13.837724685668945,
+      "learning_rate": 6.987987987987989e-06,
+      "loss": 0.5231,
+      "step": 30090
+    },
+    {
+      "epoch": 90.39,
+      "grad_norm": 9.678596496582031,
+      "learning_rate": 6.9869869869869876e-06,
+      "loss": 0.4349,
+      "step": 30100
+    },
+    {
+      "epoch": 90.42,
+      "grad_norm": 23.609298706054688,
+      "learning_rate": 6.985985985985987e-06,
+      "loss": 0.5431,
+      "step": 30110
+    },
+    {
+      "epoch": 90.45,
+      "grad_norm": 17.669897079467773,
+      "learning_rate": 6.984984984984986e-06,
+      "loss": 0.4584,
+      "step": 30120
+    },
+    {
+      "epoch": 90.48,
+      "grad_norm": 18.255775451660156,
+      "learning_rate": 6.983983983983984e-06,
+      "loss": 0.4486,
+      "step": 30130
+    },
+    {
+      "epoch": 90.51,
+      "grad_norm": 17.63195037841797,
+      "learning_rate": 6.982982982982983e-06,
+      "loss": 0.5097,
+      "step": 30140
+    },
+    {
+      "epoch": 90.54,
+      "grad_norm": 16.73892593383789,
+      "learning_rate": 6.981981981981982e-06,
+      "loss": 0.5431,
+      "step": 30150
+    },
+    {
+      "epoch": 90.57,
+      "grad_norm": 14.955709457397461,
+      "learning_rate": 6.980980980980982e-06,
+      "loss": 0.4592,
+      "step": 30160
+    },
+    {
+      "epoch": 90.6,
+      "grad_norm": 14.634506225585938,
+      "learning_rate": 6.97997997997998e-06,
+      "loss": 0.487,
+      "step": 30170
+    },
+    {
+      "epoch": 90.63,
+      "grad_norm": 13.358802795410156,
+      "learning_rate": 6.97897897897898e-06,
+      "loss": 0.4685,
+      "step": 30180
+    },
+    {
+      "epoch": 90.66,
+      "grad_norm": 12.058484077453613,
+      "learning_rate": 6.977977977977979e-06,
+      "loss": 0.5341,
+      "step": 30190
+    },
+    {
+      "epoch": 90.69,
+      "grad_norm": 12.642533302307129,
+      "learning_rate": 6.976976976976978e-06,
+      "loss": 0.5201,
+      "step": 30200
+    },
+    {
+      "epoch": 90.72,
+      "grad_norm": 12.306208610534668,
+      "learning_rate": 6.975975975975977e-06,
+      "loss": 0.4551,
+      "step": 30210
+    },
+    {
+      "epoch": 90.75,
+      "grad_norm": 17.12994384765625,
+      "learning_rate": 6.974974974974975e-06,
+      "loss": 0.5189,
+      "step": 30220
+    },
+    {
+      "epoch": 90.78,
+      "grad_norm": 14.606392860412598,
+      "learning_rate": 6.973973973973974e-06,
+      "loss": 0.509,
+      "step": 30230
+    },
+    {
+      "epoch": 90.81,
+      "grad_norm": 18.8613224029541,
+      "learning_rate": 6.972972972972973e-06,
+      "loss": 0.5104,
+      "step": 30240
+    },
+    {
+      "epoch": 90.84,
+      "grad_norm": 17.76389503479004,
+      "learning_rate": 6.971971971971972e-06,
+      "loss": 0.5088,
+      "step": 30250
+    },
+    {
+      "epoch": 90.87,
+      "grad_norm": 13.460197448730469,
+      "learning_rate": 6.970970970970972e-06,
+      "loss": 0.4629,
+      "step": 30260
+    },
+    {
+      "epoch": 90.9,
+      "grad_norm": 14.830193519592285,
+      "learning_rate": 6.9699699699699704e-06,
+      "loss": 0.5278,
+      "step": 30270
+    },
+    {
+      "epoch": 90.93,
+      "grad_norm": 12.67893123626709,
+      "learning_rate": 6.96896896896897e-06,
+      "loss": 0.5304,
+      "step": 30280
+    },
+    {
+      "epoch": 90.96,
+      "grad_norm": 17.25163459777832,
+      "learning_rate": 6.967967967967969e-06,
+      "loss": 0.5088,
+      "step": 30290
+    },
+    {
+      "epoch": 90.99,
+      "grad_norm": 21.83803939819336,
+      "learning_rate": 6.966966966966967e-06,
+      "loss": 0.4828,
+      "step": 30300
+    },
+    {
+      "epoch": 91.0,
+      "eval_accuracy": 0.8659,
+      "eval_loss": 0.4806377589702606,
+      "eval_runtime": 12.629,
+      "eval_samples_per_second": 791.831,
+      "eval_steps_per_second": 3.167,
+      "step": 30303
+    },
+    {
+      "epoch": 91.02,
+      "grad_norm": 19.721609115600586,
+      "learning_rate": 6.9659659659659665e-06,
+      "loss": 0.4452,
+      "step": 30310
+    },
+    {
+      "epoch": 91.05,
+      "grad_norm": 14.063167572021484,
+      "learning_rate": 6.964964964964965e-06,
+      "loss": 0.5178,
+      "step": 30320
+    },
+    {
+      "epoch": 91.08,
+      "grad_norm": 17.104015350341797,
+      "learning_rate": 6.9639639639639646e-06,
+      "loss": 0.4676,
+      "step": 30330
+    },
+    {
+      "epoch": 91.11,
+      "grad_norm": 16.848560333251953,
+      "learning_rate": 6.962962962962964e-06,
+      "loss": 0.4997,
+      "step": 30340
+    },
+    {
+      "epoch": 91.14,
+      "grad_norm": 15.187195777893066,
+      "learning_rate": 6.961961961961963e-06,
+      "loss": 0.483,
+      "step": 30350
+    },
+    {
+      "epoch": 91.17,
+      "grad_norm": 12.064476013183594,
+      "learning_rate": 6.960960960960962e-06,
+      "loss": 0.4512,
+      "step": 30360
+    },
+    {
+      "epoch": 91.2,
+      "grad_norm": 12.98769760131836,
+      "learning_rate": 6.959959959959961e-06,
+      "loss": 0.4519,
+      "step": 30370
+    },
+    {
+      "epoch": 91.23,
+      "grad_norm": 15.146438598632812,
+      "learning_rate": 6.958958958958959e-06,
+      "loss": 0.4664,
+      "step": 30380
+    },
+    {
+      "epoch": 91.26,
+      "grad_norm": 18.251422882080078,
+      "learning_rate": 6.957957957957958e-06,
+      "loss": 0.4693,
+      "step": 30390
+    },
+    {
+      "epoch": 91.29,
+      "grad_norm": 12.003494262695312,
+      "learning_rate": 6.956956956956957e-06,
+      "loss": 0.5161,
+      "step": 30400
+    },
+    {
+      "epoch": 91.32,
+      "grad_norm": 15.278481483459473,
+      "learning_rate": 6.955955955955957e-06,
+      "loss": 0.4667,
+      "step": 30410
+    },
+    {
+      "epoch": 91.35,
+      "grad_norm": 14.551234245300293,
+      "learning_rate": 6.954954954954955e-06,
+      "loss": 0.569,
+      "step": 30420
+    },
+    {
+      "epoch": 91.38,
+      "grad_norm": 17.031883239746094,
+      "learning_rate": 6.953953953953955e-06,
+      "loss": 0.5276,
+      "step": 30430
+    },
+    {
+      "epoch": 91.41,
+      "grad_norm": 13.356664657592773,
+      "learning_rate": 6.952952952952954e-06,
+      "loss": 0.4812,
+      "step": 30440
+    },
+    {
+      "epoch": 91.44,
+      "grad_norm": 13.456093788146973,
+      "learning_rate": 6.951951951951953e-06,
+      "loss": 0.4337,
+      "step": 30450
+    },
+    {
+      "epoch": 91.47,
+      "grad_norm": 13.668466567993164,
+      "learning_rate": 6.950950950950952e-06,
+      "loss": 0.5183,
+      "step": 30460
+    },
+    {
+      "epoch": 91.5,
+      "grad_norm": 12.544217109680176,
+      "learning_rate": 6.94994994994995e-06,
+      "loss": 0.4868,
+      "step": 30470
+    },
+    {
+      "epoch": 91.53,
+      "grad_norm": 14.428853988647461,
+      "learning_rate": 6.948948948948949e-06,
+      "loss": 0.4896,
+      "step": 30480
+    },
+    {
+      "epoch": 91.56,
+      "grad_norm": 14.298720359802246,
+      "learning_rate": 6.947947947947948e-06,
+      "loss": 0.5876,
+      "step": 30490
+    },
+    {
+      "epoch": 91.59,
+      "grad_norm": 19.13186264038086,
+      "learning_rate": 6.9469469469469474e-06,
+      "loss": 0.4919,
+      "step": 30500
+    },
+    {
+      "epoch": 91.62,
+      "grad_norm": 13.007296562194824,
+      "learning_rate": 6.945945945945947e-06,
+      "loss": 0.498,
+      "step": 30510
+    },
+    {
+      "epoch": 91.65,
+      "grad_norm": 15.038235664367676,
+      "learning_rate": 6.9449449449449455e-06,
+      "loss": 0.5064,
+      "step": 30520
+    },
+    {
+      "epoch": 91.68,
+      "grad_norm": 15.917678833007812,
+      "learning_rate": 6.943943943943945e-06,
+      "loss": 0.534,
+      "step": 30530
+    },
+    {
+      "epoch": 91.71,
+      "grad_norm": 20.900001525878906,
+      "learning_rate": 6.942942942942944e-06,
+      "loss": 0.4255,
+      "step": 30540
+    },
+    {
+      "epoch": 91.74,
+      "grad_norm": 14.348830223083496,
+      "learning_rate": 6.941941941941942e-06,
+      "loss": 0.5043,
+      "step": 30550
+    },
+    {
+      "epoch": 91.77,
+      "grad_norm": 20.730100631713867,
+      "learning_rate": 6.9409409409409416e-06,
+      "loss": 0.4839,
+      "step": 30560
+    },
+    {
+      "epoch": 91.8,
+      "grad_norm": 20.62845230102539,
+      "learning_rate": 6.93993993993994e-06,
+      "loss": 0.4621,
+      "step": 30570
+    },
+    {
+      "epoch": 91.83,
+      "grad_norm": 17.458580017089844,
+      "learning_rate": 6.93893893893894e-06,
+      "loss": 0.4559,
+      "step": 30580
+    },
+    {
+      "epoch": 91.86,
+      "grad_norm": 16.647769927978516,
+      "learning_rate": 6.937937937937938e-06,
+      "loss": 0.5066,
+      "step": 30590
+    },
+    {
+      "epoch": 91.89,
+      "grad_norm": 12.121013641357422,
+      "learning_rate": 6.936936936936938e-06,
+      "loss": 0.4387,
+      "step": 30600
+    },
+    {
+      "epoch": 91.92,
+      "grad_norm": 17.93021011352539,
+      "learning_rate": 6.935935935935937e-06,
+      "loss": 0.4736,
+      "step": 30610
+    },
+    {
+      "epoch": 91.95,
+      "grad_norm": 12.964322090148926,
+      "learning_rate": 6.934934934934936e-06,
+      "loss": 0.4667,
+      "step": 30620
+    },
+    {
+      "epoch": 91.98,
+      "grad_norm": 11.97743034362793,
+      "learning_rate": 6.933933933933934e-06,
+      "loss": 0.4357,
+      "step": 30630
+    },
+    {
+      "epoch": 92.0,
+      "eval_accuracy": 0.8676,
+      "eval_loss": 0.4770067036151886,
+      "eval_runtime": 12.9018,
+      "eval_samples_per_second": 775.083,
+      "eval_steps_per_second": 3.1,
+      "step": 30636
+    },
+    {
+      "epoch": 92.01,
+      "grad_norm": 12.249384880065918,
+      "learning_rate": 6.932932932932933e-06,
+      "loss": 0.4796,
+      "step": 30640
+    },
+    {
+      "epoch": 92.04,
+      "grad_norm": 17.532352447509766,
+      "learning_rate": 6.931931931931932e-06,
+      "loss": 0.4583,
+      "step": 30650
+    },
+    {
+      "epoch": 92.07,
+      "grad_norm": 16.517398834228516,
+      "learning_rate": 6.930930930930932e-06,
+      "loss": 0.4828,
+      "step": 30660
+    },
+    {
+      "epoch": 92.1,
+      "grad_norm": 14.836200714111328,
+      "learning_rate": 6.92992992992993e-06,
+      "loss": 0.4577,
+      "step": 30670
+    },
+    {
+      "epoch": 92.13,
+      "grad_norm": 13.0523099899292,
+      "learning_rate": 6.92892892892893e-06,
+      "loss": 0.4332,
+      "step": 30680
+    },
+    {
+      "epoch": 92.16,
+      "grad_norm": 19.937950134277344,
+      "learning_rate": 6.927927927927928e-06,
+      "loss": 0.543,
+      "step": 30690
+    },
+    {
+      "epoch": 92.19,
+      "grad_norm": 13.427345275878906,
+      "learning_rate": 6.926926926926928e-06,
+      "loss": 0.4133,
+      "step": 30700
+    },
+    {
+      "epoch": 92.22,
+      "grad_norm": 23.241262435913086,
+      "learning_rate": 6.9259259259259256e-06,
+      "loss": 0.4852,
+      "step": 30710
+    },
+    {
+      "epoch": 92.25,
+      "grad_norm": 13.879694938659668,
+      "learning_rate": 6.924924924924925e-06,
+      "loss": 0.4679,
+      "step": 30720
+    },
+    {
+      "epoch": 92.28,
+      "grad_norm": 10.753642082214355,
+      "learning_rate": 6.9239239239239244e-06,
+      "loss": 0.5127,
+      "step": 30730
+    },
+    {
+      "epoch": 92.31,
+      "grad_norm": 24.71851921081543,
+      "learning_rate": 6.922922922922923e-06,
+      "loss": 0.4803,
+      "step": 30740
+    },
+    {
+      "epoch": 92.34,
+      "grad_norm": 18.72035026550293,
+      "learning_rate": 6.9219219219219225e-06,
+      "loss": 0.4954,
+      "step": 30750
+    },
+    {
+      "epoch": 92.37,
+      "grad_norm": 13.558321952819824,
+      "learning_rate": 6.920920920920922e-06,
+      "loss": 0.4936,
+      "step": 30760
+    },
+    {
+      "epoch": 92.4,
+      "grad_norm": 16.504188537597656,
+      "learning_rate": 6.9199199199199205e-06,
+      "loss": 0.482,
+      "step": 30770
+    },
+    {
+      "epoch": 92.43,
+      "grad_norm": 20.247203826904297,
+      "learning_rate": 6.91891891891892e-06,
+      "loss": 0.5022,
+      "step": 30780
+    },
+    {
+      "epoch": 92.46,
+      "grad_norm": 10.014521598815918,
+      "learning_rate": 6.917917917917919e-06,
+      "loss": 0.4989,
+      "step": 30790
+    },
+    {
+      "epoch": 92.49,
+      "grad_norm": 10.720829010009766,
+      "learning_rate": 6.916916916916917e-06,
+      "loss": 0.4706,
+      "step": 30800
+    },
+    {
+      "epoch": 92.52,
+      "grad_norm": 20.679336547851562,
+      "learning_rate": 6.915915915915916e-06,
+      "loss": 0.4413,
+      "step": 30810
+    },
+    {
+      "epoch": 92.55,
+      "grad_norm": 17.162893295288086,
+      "learning_rate": 6.914914914914915e-06,
+      "loss": 0.5152,
+      "step": 30820
+    },
+    {
+      "epoch": 92.58,
+      "grad_norm": 13.337507247924805,
+      "learning_rate": 6.913913913913915e-06,
+      "loss": 0.4619,
+      "step": 30830
+    },
+    {
+      "epoch": 92.61,
+      "grad_norm": 16.43746566772461,
+      "learning_rate": 6.912912912912913e-06,
+      "loss": 0.5025,
+      "step": 30840
+    },
+    {
+      "epoch": 92.64,
+      "grad_norm": 16.886571884155273,
+      "learning_rate": 6.911911911911913e-06,
+      "loss": 0.4913,
+      "step": 30850
+    },
+    {
+      "epoch": 92.67,
+      "grad_norm": 16.11522102355957,
+      "learning_rate": 6.910910910910912e-06,
+      "loss": 0.4888,
+      "step": 30860
+    },
+    {
+      "epoch": 92.7,
+      "grad_norm": 19.795459747314453,
+      "learning_rate": 6.909909909909911e-06,
+      "loss": 0.5074,
+      "step": 30870
+    },
+    {
+      "epoch": 92.73,
+      "grad_norm": 16.05385398864746,
+      "learning_rate": 6.908908908908909e-06,
+      "loss": 0.4741,
+      "step": 30880
+    },
+    {
+      "epoch": 92.76,
+      "grad_norm": 19.977645874023438,
+      "learning_rate": 6.907907907907908e-06,
+      "loss": 0.4973,
+      "step": 30890
+    },
+    {
+      "epoch": 92.79,
+      "grad_norm": 14.197464942932129,
+      "learning_rate": 6.906906906906907e-06,
+      "loss": 0.5066,
+      "step": 30900
+    },
+    {
+      "epoch": 92.82,
+      "grad_norm": 15.407529830932617,
+      "learning_rate": 6.905905905905907e-06,
+      "loss": 0.465,
+      "step": 30910
+    },
+    {
+      "epoch": 92.85,
+      "grad_norm": 17.440927505493164,
+      "learning_rate": 6.904904904904905e-06,
+      "loss": 0.4783,
+      "step": 30920
+    },
+    {
+      "epoch": 92.88,
+      "grad_norm": 14.656294822692871,
+      "learning_rate": 6.903903903903905e-06,
+      "loss": 0.4681,
+      "step": 30930
+    },
+    {
+      "epoch": 92.91,
+      "grad_norm": 13.552583694458008,
+      "learning_rate": 6.902902902902903e-06,
+      "loss": 0.453,
+      "step": 30940
+    },
+    {
+      "epoch": 92.94,
+      "grad_norm": 20.369585037231445,
+      "learning_rate": 6.901901901901903e-06,
+      "loss": 0.5547,
+      "step": 30950
+    },
+    {
+      "epoch": 92.97,
+      "grad_norm": 15.291086196899414,
+      "learning_rate": 6.900900900900901e-06,
+      "loss": 0.5117,
+      "step": 30960
+    },
+    {
+      "epoch": 93.0,
+      "eval_accuracy": 0.8714,
+      "eval_loss": 0.47406187653541565,
+      "eval_runtime": 12.4967,
+      "eval_samples_per_second": 800.208,
+      "eval_steps_per_second": 3.201,
+      "step": 30969
+    },
+    {
+      "epoch": 93.0,
+      "grad_norm": 17.182878494262695,
+      "learning_rate": 6.8998998998999e-06,
+      "loss": 0.4386,
+      "step": 30970
+    },
+    {
+      "epoch": 93.03,
+      "grad_norm": 13.908241271972656,
+      "learning_rate": 6.8988988988988995e-06,
+      "loss": 0.4534,
+      "step": 30980
+    },
+    {
+      "epoch": 93.06,
+      "grad_norm": 14.219406127929688,
+      "learning_rate": 6.897897897897898e-06,
+      "loss": 0.4639,
+      "step": 30990
+    },
+    {
+      "epoch": 93.09,
+      "grad_norm": 21.001972198486328,
+      "learning_rate": 6.8968968968968975e-06,
+      "loss": 0.482,
+      "step": 31000
+    },
+    {
+      "epoch": 93.12,
+      "grad_norm": 15.324588775634766,
+      "learning_rate": 6.895895895895897e-06,
+      "loss": 0.5158,
+      "step": 31010
+    },
+    {
+      "epoch": 93.15,
+      "grad_norm": 12.66370677947998,
+      "learning_rate": 6.8948948948948955e-06,
+      "loss": 0.4488,
+      "step": 31020
+    },
+    {
+      "epoch": 93.18,
+      "grad_norm": 12.25474739074707,
+      "learning_rate": 6.893893893893895e-06,
+      "loss": 0.4838,
+      "step": 31030
+    },
+    {
+      "epoch": 93.21,
+      "grad_norm": 14.16290283203125,
+      "learning_rate": 6.892892892892894e-06,
+      "loss": 0.4867,
+      "step": 31040
+    },
+    {
+      "epoch": 93.24,
+      "grad_norm": 18.15909194946289,
+      "learning_rate": 6.891891891891892e-06,
+      "loss": 0.5054,
+      "step": 31050
+    },
+    {
+      "epoch": 93.27,
+      "grad_norm": 22.629234313964844,
+      "learning_rate": 6.890890890890891e-06,
+      "loss": 0.531,
+      "step": 31060
+    },
+    {
+      "epoch": 93.3,
+      "grad_norm": 14.281317710876465,
+      "learning_rate": 6.88988988988989e-06,
+      "loss": 0.4178,
+      "step": 31070
+    },
+    {
+      "epoch": 93.33,
+      "grad_norm": 11.705520629882812,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 0.4821,
+      "step": 31080
+    },
+    {
+      "epoch": 93.36,
+      "grad_norm": 15.612814903259277,
+      "learning_rate": 6.887887887887888e-06,
+      "loss": 0.4601,
+      "step": 31090
+    },
+    {
+      "epoch": 93.39,
+      "grad_norm": 15.573897361755371,
+      "learning_rate": 6.886886886886888e-06,
+      "loss": 0.4857,
+      "step": 31100
+    },
+    {
+      "epoch": 93.42,
+      "grad_norm": 18.67922019958496,
+      "learning_rate": 6.885885885885887e-06,
+      "loss": 0.52,
+      "step": 31110
+    },
+    {
+      "epoch": 93.45,
+      "grad_norm": 17.691247940063477,
+      "learning_rate": 6.884884884884886e-06,
+      "loss": 0.4484,
+      "step": 31120
+    },
+    {
+      "epoch": 93.48,
+      "grad_norm": 19.116165161132812,
+      "learning_rate": 6.883883883883884e-06,
+      "loss": 0.4609,
+      "step": 31130
+    },
+    {
+      "epoch": 93.51,
+      "grad_norm": 22.42650032043457,
+      "learning_rate": 6.882882882882883e-06,
+      "loss": 0.4463,
+      "step": 31140
+    },
+    {
+      "epoch": 93.54,
+      "grad_norm": 15.270142555236816,
+      "learning_rate": 6.881881881881882e-06,
+      "loss": 0.518,
+      "step": 31150
+    },
+    {
+      "epoch": 93.57,
+      "grad_norm": 20.749954223632812,
+      "learning_rate": 6.880880880880881e-06,
+      "loss": 0.5692,
+      "step": 31160
+    },
+    {
+      "epoch": 93.6,
+      "grad_norm": 12.274581909179688,
+      "learning_rate": 6.87987987987988e-06,
+      "loss": 0.4486,
+      "step": 31170
+    },
+    {
+      "epoch": 93.63,
+      "grad_norm": 15.650559425354004,
+      "learning_rate": 6.87887887887888e-06,
+      "loss": 0.5417,
+      "step": 31180
+    },
+    {
+      "epoch": 93.66,
+      "grad_norm": 20.807689666748047,
+      "learning_rate": 6.8778778778778784e-06,
+      "loss": 0.5153,
+      "step": 31190
+    },
+    {
+      "epoch": 93.69,
+      "grad_norm": 16.74406623840332,
+      "learning_rate": 6.876876876876878e-06,
+      "loss": 0.5254,
+      "step": 31200
+    },
+    {
+      "epoch": 93.72,
+      "grad_norm": 14.476408004760742,
+      "learning_rate": 6.875875875875876e-06,
+      "loss": 0.46,
+      "step": 31210
+    },
+    {
+      "epoch": 93.75,
+      "grad_norm": 14.962093353271484,
+      "learning_rate": 6.874874874874875e-06,
+      "loss": 0.4943,
+      "step": 31220
+    },
+    {
+      "epoch": 93.78,
+      "grad_norm": 13.664773941040039,
+      "learning_rate": 6.8738738738738745e-06,
+      "loss": 0.5187,
+      "step": 31230
+    },
+    {
+      "epoch": 93.81,
+      "grad_norm": 19.524948120117188,
+      "learning_rate": 6.872872872872873e-06,
+      "loss": 0.4744,
+      "step": 31240
+    },
+    {
+      "epoch": 93.84,
+      "grad_norm": 11.793402671813965,
+      "learning_rate": 6.8718718718718725e-06,
+      "loss": 0.5423,
+      "step": 31250
+    },
+    {
+      "epoch": 93.87,
+      "grad_norm": 13.440789222717285,
+      "learning_rate": 6.870870870870872e-06,
+      "loss": 0.4325,
+      "step": 31260
+    },
+    {
+      "epoch": 93.9,
+      "grad_norm": 18.27890968322754,
+      "learning_rate": 6.869869869869871e-06,
+      "loss": 0.4928,
+      "step": 31270
+    },
+    {
+      "epoch": 93.93,
+      "grad_norm": 14.388352394104004,
+      "learning_rate": 6.86886886886887e-06,
+      "loss": 0.5074,
+      "step": 31280
+    },
+    {
+      "epoch": 93.96,
+      "grad_norm": 13.075735092163086,
+      "learning_rate": 6.867867867867869e-06,
+      "loss": 0.4386,
+      "step": 31290
+    },
+    {
+      "epoch": 93.99,
+      "grad_norm": 14.853118896484375,
+      "learning_rate": 6.866866866866867e-06,
+      "loss": 0.4756,
+      "step": 31300
+    },
+    {
+      "epoch": 94.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.48601529002189636,
+      "eval_runtime": 13.0414,
+      "eval_samples_per_second": 766.79,
+      "eval_steps_per_second": 3.067,
+      "step": 31302
+    },
+    {
+      "epoch": 94.02,
+      "grad_norm": 18.53443717956543,
+      "learning_rate": 6.865865865865866e-06,
+      "loss": 0.4487,
+      "step": 31310
+    },
+    {
+      "epoch": 94.05,
+      "grad_norm": 21.4047794342041,
+      "learning_rate": 6.864864864864865e-06,
+      "loss": 0.4984,
+      "step": 31320
+    },
+    {
+      "epoch": 94.08,
+      "grad_norm": 18.063899993896484,
+      "learning_rate": 6.863863863863865e-06,
+      "loss": 0.532,
+      "step": 31330
+    },
+    {
+      "epoch": 94.11,
+      "grad_norm": 18.434669494628906,
+      "learning_rate": 6.862862862862863e-06,
+      "loss": 0.5251,
+      "step": 31340
+    },
+    {
+      "epoch": 94.14,
+      "grad_norm": 14.309115409851074,
+      "learning_rate": 6.861861861861863e-06,
+      "loss": 0.4703,
+      "step": 31350
+    },
+    {
+      "epoch": 94.17,
+      "grad_norm": 15.579237937927246,
+      "learning_rate": 6.860860860860862e-06,
+      "loss": 0.4788,
+      "step": 31360
+    },
+    {
+      "epoch": 94.2,
+      "grad_norm": 12.798219680786133,
+      "learning_rate": 6.859859859859861e-06,
+      "loss": 0.4855,
+      "step": 31370
+    },
+    {
+      "epoch": 94.23,
+      "grad_norm": 10.201995849609375,
+      "learning_rate": 6.8588588588588585e-06,
+      "loss": 0.4595,
+      "step": 31380
+    },
+    {
+      "epoch": 94.26,
+      "grad_norm": 16.886245727539062,
+      "learning_rate": 6.857857857857858e-06,
+      "loss": 0.4965,
+      "step": 31390
+    },
+    {
+      "epoch": 94.29,
+      "grad_norm": 17.06256866455078,
+      "learning_rate": 6.856856856856857e-06,
+      "loss": 0.5013,
+      "step": 31400
+    },
+    {
+      "epoch": 94.32,
+      "grad_norm": 15.504446983337402,
+      "learning_rate": 6.855855855855856e-06,
+      "loss": 0.4848,
+      "step": 31410
+    },
+    {
+      "epoch": 94.35,
+      "grad_norm": 31.186729431152344,
+      "learning_rate": 6.8548548548548554e-06,
+      "loss": 0.4897,
+      "step": 31420
+    },
+    {
+      "epoch": 94.38,
+      "grad_norm": 16.205385208129883,
+      "learning_rate": 6.853853853853855e-06,
+      "loss": 0.4479,
+      "step": 31430
+    },
+    {
+      "epoch": 94.41,
+      "grad_norm": 18.163265228271484,
+      "learning_rate": 6.8528528528528535e-06,
+      "loss": 0.4501,
+      "step": 31440
+    },
+    {
+      "epoch": 94.44,
+      "grad_norm": 16.375808715820312,
+      "learning_rate": 6.851851851851853e-06,
+      "loss": 0.4955,
+      "step": 31450
+    },
+    {
+      "epoch": 94.47,
+      "grad_norm": 12.83496379852295,
+      "learning_rate": 6.850850850850851e-06,
+      "loss": 0.4409,
+      "step": 31460
+    },
+    {
+      "epoch": 94.5,
+      "grad_norm": 19.523500442504883,
+      "learning_rate": 6.84984984984985e-06,
+      "loss": 0.5423,
+      "step": 31470
+    },
+    {
+      "epoch": 94.53,
+      "grad_norm": 27.643802642822266,
+      "learning_rate": 6.8488488488488495e-06,
+      "loss": 0.5097,
+      "step": 31480
+    },
+    {
+      "epoch": 94.56,
+      "grad_norm": 18.769577026367188,
+      "learning_rate": 6.847847847847848e-06,
+      "loss": 0.513,
+      "step": 31490
+    },
+    {
+      "epoch": 94.59,
+      "grad_norm": 15.564085006713867,
+      "learning_rate": 6.846846846846848e-06,
+      "loss": 0.5179,
+      "step": 31500
+    },
+    {
+      "epoch": 94.62,
+      "grad_norm": 13.962645530700684,
+      "learning_rate": 6.845845845845846e-06,
+      "loss": 0.4604,
+      "step": 31510
+    },
+    {
+      "epoch": 94.65,
+      "grad_norm": 13.918479919433594,
+      "learning_rate": 6.844844844844846e-06,
+      "loss": 0.4892,
+      "step": 31520
+    },
+    {
+      "epoch": 94.68,
+      "grad_norm": 15.171954154968262,
+      "learning_rate": 6.843843843843845e-06,
+      "loss": 0.4332,
+      "step": 31530
+    },
+    {
+      "epoch": 94.71,
+      "grad_norm": 18.3940486907959,
+      "learning_rate": 6.842842842842844e-06,
+      "loss": 0.4625,
+      "step": 31540
+    },
+    {
+      "epoch": 94.74,
+      "grad_norm": 15.39448356628418,
+      "learning_rate": 6.841841841841842e-06,
+      "loss": 0.5074,
+      "step": 31550
+    },
+    {
+      "epoch": 94.77,
+      "grad_norm": 19.629560470581055,
+      "learning_rate": 6.840840840840841e-06,
+      "loss": 0.462,
+      "step": 31560
+    },
+    {
+      "epoch": 94.8,
+      "grad_norm": 23.34688377380371,
+      "learning_rate": 6.83983983983984e-06,
+      "loss": 0.4677,
+      "step": 31570
+    },
+    {
+      "epoch": 94.83,
+      "grad_norm": 19.711841583251953,
+      "learning_rate": 6.83883883883884e-06,
+      "loss": 0.478,
+      "step": 31580
+    },
+    {
+      "epoch": 94.86,
+      "grad_norm": 20.41490364074707,
+      "learning_rate": 6.837837837837838e-06,
+      "loss": 0.4859,
+      "step": 31590
+    },
+    {
+      "epoch": 94.89,
+      "grad_norm": 14.030795097351074,
+      "learning_rate": 6.836836836836838e-06,
+      "loss": 0.5007,
+      "step": 31600
+    },
+    {
+      "epoch": 94.92,
+      "grad_norm": 26.083614349365234,
+      "learning_rate": 6.835835835835836e-06,
+      "loss": 0.4975,
+      "step": 31610
+    },
+    {
+      "epoch": 94.95,
+      "grad_norm": 13.442200660705566,
+      "learning_rate": 6.834834834834836e-06,
+      "loss": 0.4878,
+      "step": 31620
+    },
+    {
+      "epoch": 94.98,
+      "grad_norm": 16.380163192749023,
+      "learning_rate": 6.8338338338338336e-06,
+      "loss": 0.4575,
+      "step": 31630
+    },
+    {
+      "epoch": 95.0,
+      "eval_accuracy": 0.8652,
+      "eval_loss": 0.4854680597782135,
+      "eval_runtime": 12.937,
+      "eval_samples_per_second": 772.979,
+      "eval_steps_per_second": 3.092,
+      "step": 31635
+    },
+    {
+      "epoch": 95.02,
+      "grad_norm": 19.621206283569336,
+      "learning_rate": 6.832832832832833e-06,
+      "loss": 0.4757,
+      "step": 31640
+    },
+    {
+      "epoch": 95.05,
+      "grad_norm": 13.114680290222168,
+      "learning_rate": 6.8318318318318324e-06,
+      "loss": 0.4251,
+      "step": 31650
+    },
+    {
+      "epoch": 95.08,
+      "grad_norm": 21.07643699645996,
+      "learning_rate": 6.830830830830831e-06,
+      "loss": 0.4906,
+      "step": 31660
+    },
+    {
+      "epoch": 95.11,
+      "grad_norm": 16.28285789489746,
+      "learning_rate": 6.8298298298298305e-06,
+      "loss": 0.453,
+      "step": 31670
+    },
+    {
+      "epoch": 95.14,
+      "grad_norm": 15.50754165649414,
+      "learning_rate": 6.82882882882883e-06,
+      "loss": 0.4508,
+      "step": 31680
+    },
+    {
+      "epoch": 95.17,
+      "grad_norm": 16.43500328063965,
+      "learning_rate": 6.8278278278278285e-06,
+      "loss": 0.4899,
+      "step": 31690
+    },
+    {
+      "epoch": 95.2,
+      "grad_norm": 23.96570587158203,
+      "learning_rate": 6.826826826826828e-06,
+      "loss": 0.5045,
+      "step": 31700
+    },
+    {
+      "epoch": 95.23,
+      "grad_norm": 12.40204906463623,
+      "learning_rate": 6.825825825825826e-06,
+      "loss": 0.5175,
+      "step": 31710
+    },
+    {
+      "epoch": 95.26,
+      "grad_norm": 19.6379451751709,
+      "learning_rate": 6.824824824824825e-06,
+      "loss": 0.4824,
+      "step": 31720
+    },
+    {
+      "epoch": 95.29,
+      "grad_norm": 15.54357624053955,
+      "learning_rate": 6.823823823823824e-06,
+      "loss": 0.5164,
+      "step": 31730
+    },
+    {
+      "epoch": 95.32,
+      "grad_norm": 23.734405517578125,
+      "learning_rate": 6.822822822822823e-06,
+      "loss": 0.4908,
+      "step": 31740
+    },
+    {
+      "epoch": 95.35,
+      "grad_norm": 14.812207221984863,
+      "learning_rate": 6.821821821821823e-06,
+      "loss": 0.4388,
+      "step": 31750
+    },
+    {
+      "epoch": 95.38,
+      "grad_norm": 22.30416488647461,
+      "learning_rate": 6.820820820820821e-06,
+      "loss": 0.5026,
+      "step": 31760
+    },
+    {
+      "epoch": 95.41,
+      "grad_norm": 15.008042335510254,
+      "learning_rate": 6.819819819819821e-06,
+      "loss": 0.4348,
+      "step": 31770
+    },
+    {
+      "epoch": 95.44,
+      "grad_norm": 28.234743118286133,
+      "learning_rate": 6.81881881881882e-06,
+      "loss": 0.4816,
+      "step": 31780
+    },
+    {
+      "epoch": 95.47,
+      "grad_norm": 13.401700019836426,
+      "learning_rate": 6.817817817817819e-06,
+      "loss": 0.4602,
+      "step": 31790
+    },
+    {
+      "epoch": 95.5,
+      "grad_norm": 16.137550354003906,
+      "learning_rate": 6.816816816816817e-06,
+      "loss": 0.4929,
+      "step": 31800
+    },
+    {
+      "epoch": 95.53,
+      "grad_norm": 11.09272289276123,
+      "learning_rate": 6.815815815815816e-06,
+      "loss": 0.4779,
+      "step": 31810
+    },
+    {
+      "epoch": 95.56,
+      "grad_norm": 14.058930397033691,
+      "learning_rate": 6.814814814814815e-06,
+      "loss": 0.4599,
+      "step": 31820
+    },
+    {
+      "epoch": 95.59,
+      "grad_norm": 18.617870330810547,
+      "learning_rate": 6.813813813813815e-06,
+      "loss": 0.4796,
+      "step": 31830
+    },
+    {
+      "epoch": 95.62,
+      "grad_norm": 15.510306358337402,
+      "learning_rate": 6.812812812812813e-06,
+      "loss": 0.4697,
+      "step": 31840
+    },
+    {
+      "epoch": 95.65,
+      "grad_norm": 18.052316665649414,
+      "learning_rate": 6.811811811811813e-06,
+      "loss": 0.4792,
+      "step": 31850
+    },
+    {
+      "epoch": 95.68,
+      "grad_norm": 15.386200904846191,
+      "learning_rate": 6.810810810810811e-06,
+      "loss": 0.4866,
+      "step": 31860
+    },
+    {
+      "epoch": 95.71,
+      "grad_norm": 16.352357864379883,
+      "learning_rate": 6.809809809809811e-06,
+      "loss": 0.5187,
+      "step": 31870
+    },
+    {
+      "epoch": 95.74,
+      "grad_norm": 12.986818313598633,
+      "learning_rate": 6.808808808808809e-06,
+      "loss": 0.4311,
+      "step": 31880
+    },
+    {
+      "epoch": 95.77,
+      "grad_norm": 12.23602294921875,
+      "learning_rate": 6.807807807807808e-06,
+      "loss": 0.4361,
+      "step": 31890
+    },
+    {
+      "epoch": 95.8,
+      "grad_norm": 12.55474853515625,
+      "learning_rate": 6.8068068068068075e-06,
+      "loss": 0.5035,
+      "step": 31900
+    },
+    {
+      "epoch": 95.83,
+      "grad_norm": 23.33086585998535,
+      "learning_rate": 6.805805805805806e-06,
+      "loss": 0.4445,
+      "step": 31910
+    },
+    {
+      "epoch": 95.86,
+      "grad_norm": 16.66939926147461,
+      "learning_rate": 6.8048048048048055e-06,
+      "loss": 0.495,
+      "step": 31920
+    },
+    {
+      "epoch": 95.89,
+      "grad_norm": 14.644975662231445,
+      "learning_rate": 6.803803803803805e-06,
+      "loss": 0.5221,
+      "step": 31930
+    },
+    {
+      "epoch": 95.92,
+      "grad_norm": 11.44594955444336,
+      "learning_rate": 6.8028028028028035e-06,
+      "loss": 0.4813,
+      "step": 31940
+    },
+    {
+      "epoch": 95.95,
+      "grad_norm": 17.599153518676758,
+      "learning_rate": 6.801801801801803e-06,
+      "loss": 0.4931,
+      "step": 31950
+    },
+    {
+      "epoch": 95.98,
+      "grad_norm": 15.345081329345703,
+      "learning_rate": 6.800800800800801e-06,
+      "loss": 0.4657,
+      "step": 31960
+    },
+    {
+      "epoch": 96.0,
+      "eval_accuracy": 0.8677,
+      "eval_loss": 0.4828157126903534,
+      "eval_runtime": 12.5806,
+      "eval_samples_per_second": 794.875,
+      "eval_steps_per_second": 3.18,
+      "step": 31968
+    },
+    {
+      "epoch": 96.01,
+      "grad_norm": 15.16838264465332,
+      "learning_rate": 6.7997997997998e-06,
+      "loss": 0.4298,
+      "step": 31970
+    },
+    {
+      "epoch": 96.04,
+      "grad_norm": 19.645170211791992,
+      "learning_rate": 6.798798798798799e-06,
+      "loss": 0.4518,
+      "step": 31980
+    },
+    {
+      "epoch": 96.07,
+      "grad_norm": 21.45229148864746,
+      "learning_rate": 6.797797797797798e-06,
+      "loss": 0.4612,
+      "step": 31990
+    },
+    {
+      "epoch": 96.1,
+      "grad_norm": 15.509103775024414,
+      "learning_rate": 6.796796796796798e-06,
+      "loss": 0.4415,
+      "step": 32000
+    },
+    {
+      "epoch": 96.13,
+      "grad_norm": 14.290263175964355,
+      "learning_rate": 6.795795795795796e-06,
+      "loss": 0.4475,
+      "step": 32010
+    },
+    {
+      "epoch": 96.16,
+      "grad_norm": 13.262653350830078,
+      "learning_rate": 6.794794794794796e-06,
+      "loss": 0.4582,
+      "step": 32020
+    },
+    {
+      "epoch": 96.19,
+      "grad_norm": 11.732439994812012,
+      "learning_rate": 6.793793793793795e-06,
+      "loss": 0.4829,
+      "step": 32030
+    },
+    {
+      "epoch": 96.22,
+      "grad_norm": 22.722179412841797,
+      "learning_rate": 6.792792792792793e-06,
+      "loss": 0.5062,
+      "step": 32040
+    },
+    {
+      "epoch": 96.25,
+      "grad_norm": 15.500922203063965,
+      "learning_rate": 6.791791791791792e-06,
+      "loss": 0.4561,
+      "step": 32050
+    },
+    {
+      "epoch": 96.28,
+      "grad_norm": 12.340556144714355,
+      "learning_rate": 6.790790790790791e-06,
+      "loss": 0.4458,
+      "step": 32060
+    },
+    {
+      "epoch": 96.31,
+      "grad_norm": 10.26012897491455,
+      "learning_rate": 6.78978978978979e-06,
+      "loss": 0.5053,
+      "step": 32070
+    },
+    {
+      "epoch": 96.34,
+      "grad_norm": 18.758434295654297,
+      "learning_rate": 6.788788788788789e-06,
+      "loss": 0.5106,
+      "step": 32080
+    },
+    {
+      "epoch": 96.37,
+      "grad_norm": 13.208051681518555,
+      "learning_rate": 6.787787787787788e-06,
+      "loss": 0.5009,
+      "step": 32090
+    },
+    {
+      "epoch": 96.4,
+      "grad_norm": 10.817633628845215,
+      "learning_rate": 6.786786786786788e-06,
+      "loss": 0.4505,
+      "step": 32100
+    },
+    {
+      "epoch": 96.43,
+      "grad_norm": 15.572339057922363,
+      "learning_rate": 6.7857857857857864e-06,
+      "loss": 0.4375,
+      "step": 32110
+    },
+    {
+      "epoch": 96.46,
+      "grad_norm": 11.524653434753418,
+      "learning_rate": 6.784784784784786e-06,
+      "loss": 0.4877,
+      "step": 32120
+    },
+    {
+      "epoch": 96.49,
+      "grad_norm": 10.657195091247559,
+      "learning_rate": 6.783783783783784e-06,
+      "loss": 0.4458,
+      "step": 32130
+    },
+    {
+      "epoch": 96.52,
+      "grad_norm": 23.2871150970459,
+      "learning_rate": 6.782782782782783e-06,
+      "loss": 0.4696,
+      "step": 32140
+    },
+    {
+      "epoch": 96.55,
+      "grad_norm": 22.009185791015625,
+      "learning_rate": 6.7817817817817825e-06,
+      "loss": 0.4779,
+      "step": 32150
+    },
+    {
+      "epoch": 96.58,
+      "grad_norm": 12.608153343200684,
+      "learning_rate": 6.780780780780781e-06,
+      "loss": 0.4377,
+      "step": 32160
+    },
+    {
+      "epoch": 96.61,
+      "grad_norm": 18.28102684020996,
+      "learning_rate": 6.7797797797797805e-06,
+      "loss": 0.5006,
+      "step": 32170
+    },
+    {
+      "epoch": 96.64,
+      "grad_norm": 14.66008472442627,
+      "learning_rate": 6.778778778778779e-06,
+      "loss": 0.5115,
+      "step": 32180
+    },
+    {
+      "epoch": 96.67,
+      "grad_norm": 21.086063385009766,
+      "learning_rate": 6.777777777777779e-06,
+      "loss": 0.4947,
+      "step": 32190
+    },
+    {
+      "epoch": 96.7,
+      "grad_norm": 16.663759231567383,
+      "learning_rate": 6.776776776776778e-06,
+      "loss": 0.5101,
+      "step": 32200
+    },
+    {
+      "epoch": 96.73,
+      "grad_norm": 16.673625946044922,
+      "learning_rate": 6.775775775775776e-06,
+      "loss": 0.4977,
+      "step": 32210
+    },
+    {
+      "epoch": 96.76,
+      "grad_norm": 12.190402030944824,
+      "learning_rate": 6.774774774774775e-06,
+      "loss": 0.4597,
+      "step": 32220
+    },
+    {
+      "epoch": 96.79,
+      "grad_norm": 16.515329360961914,
+      "learning_rate": 6.773773773773774e-06,
+      "loss": 0.4762,
+      "step": 32230
+    },
+    {
+      "epoch": 96.82,
+      "grad_norm": 18.57062530517578,
+      "learning_rate": 6.772772772772773e-06,
+      "loss": 0.4923,
+      "step": 32240
+    },
+    {
+      "epoch": 96.85,
+      "grad_norm": 18.67723846435547,
+      "learning_rate": 6.771771771771773e-06,
+      "loss": 0.4464,
+      "step": 32250
+    },
+    {
+      "epoch": 96.88,
+      "grad_norm": 15.1109619140625,
+      "learning_rate": 6.770770770770771e-06,
+      "loss": 0.4803,
+      "step": 32260
+    },
+    {
+      "epoch": 96.91,
+      "grad_norm": 11.679488182067871,
+      "learning_rate": 6.769769769769771e-06,
+      "loss": 0.4442,
+      "step": 32270
+    },
+    {
+      "epoch": 96.94,
+      "grad_norm": 21.143169403076172,
+      "learning_rate": 6.76876876876877e-06,
+      "loss": 0.4501,
+      "step": 32280
+    },
+    {
+      "epoch": 96.97,
+      "grad_norm": 11.951123237609863,
+      "learning_rate": 6.767767767767768e-06,
+      "loss": 0.4627,
+      "step": 32290
+    },
+    {
+      "epoch": 97.0,
+      "grad_norm": 17.269346237182617,
+      "learning_rate": 6.7667667667667665e-06,
+      "loss": 0.4746,
+      "step": 32300
+    },
+    {
+      "epoch": 97.0,
+      "eval_accuracy": 0.8676,
+      "eval_loss": 0.48496025800704956,
+      "eval_runtime": 12.6089,
+      "eval_samples_per_second": 793.09,
+      "eval_steps_per_second": 3.172,
+      "step": 32301
+    },
+    {
+      "epoch": 97.03,
+      "grad_norm": 12.833151817321777,
+      "learning_rate": 6.765765765765766e-06,
+      "loss": 0.5508,
+      "step": 32310
+    },
+    {
+      "epoch": 97.06,
+      "grad_norm": 18.792051315307617,
+      "learning_rate": 6.764764764764765e-06,
+      "loss": 0.4696,
+      "step": 32320
+    },
+    {
+      "epoch": 97.09,
+      "grad_norm": 14.260638236999512,
+      "learning_rate": 6.763763763763764e-06,
+      "loss": 0.4844,
+      "step": 32330
+    },
+    {
+      "epoch": 97.12,
+      "grad_norm": 12.716498374938965,
+      "learning_rate": 6.7627627627627634e-06,
+      "loss": 0.4656,
+      "step": 32340
+    },
+    {
+      "epoch": 97.15,
+      "grad_norm": 17.392988204956055,
+      "learning_rate": 6.761761761761763e-06,
+      "loss": 0.3913,
+      "step": 32350
+    },
+    {
+      "epoch": 97.18,
+      "grad_norm": 20.23369026184082,
+      "learning_rate": 6.7607607607607615e-06,
+      "loss": 0.4892,
+      "step": 32360
+    },
+    {
+      "epoch": 97.21,
+      "grad_norm": 24.130441665649414,
+      "learning_rate": 6.759759759759761e-06,
+      "loss": 0.5021,
+      "step": 32370
+    },
+    {
+      "epoch": 97.24,
+      "grad_norm": 11.284120559692383,
+      "learning_rate": 6.758758758758759e-06,
+      "loss": 0.4619,
+      "step": 32380
+    },
+    {
+      "epoch": 97.27,
+      "grad_norm": 15.775595664978027,
+      "learning_rate": 6.757757757757758e-06,
+      "loss": 0.471,
+      "step": 32390
+    },
+    {
+      "epoch": 97.3,
+      "grad_norm": 20.100561141967773,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.4674,
+      "step": 32400
+    },
+    {
+      "epoch": 97.33,
+      "grad_norm": 12.14714527130127,
+      "learning_rate": 6.755755755755756e-06,
+      "loss": 0.5125,
+      "step": 32410
+    },
+    {
+      "epoch": 97.36,
+      "grad_norm": 24.397743225097656,
+      "learning_rate": 6.7547547547547556e-06,
+      "loss": 0.4616,
+      "step": 32420
+    },
+    {
+      "epoch": 97.39,
+      "grad_norm": 13.80153751373291,
+      "learning_rate": 6.753753753753754e-06,
+      "loss": 0.4486,
+      "step": 32430
+    },
+    {
+      "epoch": 97.42,
+      "grad_norm": 15.475515365600586,
+      "learning_rate": 6.752752752752754e-06,
+      "loss": 0.4943,
+      "step": 32440
+    },
+    {
+      "epoch": 97.45,
+      "grad_norm": 13.154668807983398,
+      "learning_rate": 6.751751751751753e-06,
+      "loss": 0.5027,
+      "step": 32450
+    },
+    {
+      "epoch": 97.48,
+      "grad_norm": 16.03565216064453,
+      "learning_rate": 6.750750750750751e-06,
+      "loss": 0.4424,
+      "step": 32460
+    },
+    {
+      "epoch": 97.51,
+      "grad_norm": 11.592856407165527,
+      "learning_rate": 6.74974974974975e-06,
+      "loss": 0.504,
+      "step": 32470
+    },
+    {
+      "epoch": 97.54,
+      "grad_norm": 14.62682056427002,
+      "learning_rate": 6.748748748748749e-06,
+      "loss": 0.4428,
+      "step": 32480
+    },
+    {
+      "epoch": 97.57,
+      "grad_norm": 20.040401458740234,
+      "learning_rate": 6.747747747747748e-06,
+      "loss": 0.4604,
+      "step": 32490
+    },
+    {
+      "epoch": 97.6,
+      "grad_norm": 13.525559425354004,
+      "learning_rate": 6.746746746746748e-06,
+      "loss": 0.5065,
+      "step": 32500
+    },
+    {
+      "epoch": 97.63,
+      "grad_norm": 19.53817367553711,
+      "learning_rate": 6.745745745745746e-06,
+      "loss": 0.4831,
+      "step": 32510
+    },
+    {
+      "epoch": 97.66,
+      "grad_norm": 16.614377975463867,
+      "learning_rate": 6.744744744744746e-06,
+      "loss": 0.435,
+      "step": 32520
+    },
+    {
+      "epoch": 97.69,
+      "grad_norm": 16.076602935791016,
+      "learning_rate": 6.743743743743744e-06,
+      "loss": 0.4448,
+      "step": 32530
+    },
+    {
+      "epoch": 97.72,
+      "grad_norm": 16.50364875793457,
+      "learning_rate": 6.742742742742743e-06,
+      "loss": 0.475,
+      "step": 32540
+    },
+    {
+      "epoch": 97.75,
+      "grad_norm": 20.241161346435547,
+      "learning_rate": 6.7417417417417415e-06,
+      "loss": 0.4473,
+      "step": 32550
+    },
+    {
+      "epoch": 97.78,
+      "grad_norm": 12.64549446105957,
+      "learning_rate": 6.740740740740741e-06,
+      "loss": 0.4589,
+      "step": 32560
+    },
+    {
+      "epoch": 97.81,
+      "grad_norm": 14.082094192504883,
+      "learning_rate": 6.7397397397397404e-06,
+      "loss": 0.4763,
+      "step": 32570
+    },
+    {
+      "epoch": 97.84,
+      "grad_norm": 18.26226806640625,
+      "learning_rate": 6.738738738738739e-06,
+      "loss": 0.4076,
+      "step": 32580
+    },
+    {
+      "epoch": 97.87,
+      "grad_norm": 10.316207885742188,
+      "learning_rate": 6.7377377377377385e-06,
+      "loss": 0.4297,
+      "step": 32590
+    },
+    {
+      "epoch": 97.9,
+      "grad_norm": 20.827877044677734,
+      "learning_rate": 6.736736736736738e-06,
+      "loss": 0.4637,
+      "step": 32600
+    },
+    {
+      "epoch": 97.93,
+      "grad_norm": 13.130403518676758,
+      "learning_rate": 6.7357357357357365e-06,
+      "loss": 0.5148,
+      "step": 32610
+    },
+    {
+      "epoch": 97.96,
+      "grad_norm": 19.18124771118164,
+      "learning_rate": 6.734734734734736e-06,
+      "loss": 0.508,
+      "step": 32620
+    },
+    {
+      "epoch": 97.99,
+      "grad_norm": 14.950023651123047,
+      "learning_rate": 6.733733733733734e-06,
+      "loss": 0.5466,
+      "step": 32630
+    },
+    {
+      "epoch": 98.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.48899784684181213,
+      "eval_runtime": 13.0523,
+      "eval_samples_per_second": 766.149,
+      "eval_steps_per_second": 3.065,
+      "step": 32634
+    },
+    {
+      "epoch": 98.02,
+      "grad_norm": 18.302017211914062,
+      "learning_rate": 6.732732732732733e-06,
+      "loss": 0.5146,
+      "step": 32640
+    },
+    {
+      "epoch": 98.05,
+      "grad_norm": 13.174702644348145,
+      "learning_rate": 6.731731731731732e-06,
+      "loss": 0.5268,
+      "step": 32650
+    },
+    {
+      "epoch": 98.08,
+      "grad_norm": 14.635217666625977,
+      "learning_rate": 6.730730730730731e-06,
+      "loss": 0.4408,
+      "step": 32660
+    },
+    {
+      "epoch": 98.11,
+      "grad_norm": 17.96314239501953,
+      "learning_rate": 6.729729729729731e-06,
+      "loss": 0.4223,
+      "step": 32670
+    },
+    {
+      "epoch": 98.14,
+      "grad_norm": 18.2534236907959,
+      "learning_rate": 6.728728728728729e-06,
+      "loss": 0.5546,
+      "step": 32680
+    },
+    {
+      "epoch": 98.17,
+      "grad_norm": 10.7394380569458,
+      "learning_rate": 6.727727727727729e-06,
+      "loss": 0.5166,
+      "step": 32690
+    },
+    {
+      "epoch": 98.2,
+      "grad_norm": 13.858802795410156,
+      "learning_rate": 6.726726726726728e-06,
+      "loss": 0.449,
+      "step": 32700
+    },
+    {
+      "epoch": 98.23,
+      "grad_norm": 16.374286651611328,
+      "learning_rate": 6.725725725725726e-06,
+      "loss": 0.4717,
+      "step": 32710
+    },
+    {
+      "epoch": 98.26,
+      "grad_norm": 10.098102569580078,
+      "learning_rate": 6.724724724724725e-06,
+      "loss": 0.5092,
+      "step": 32720
+    },
+    {
+      "epoch": 98.29,
+      "grad_norm": 18.6304874420166,
+      "learning_rate": 6.723723723723724e-06,
+      "loss": 0.4877,
+      "step": 32730
+    },
+    {
+      "epoch": 98.32,
+      "grad_norm": 14.151505470275879,
+      "learning_rate": 6.722722722722723e-06,
+      "loss": 0.468,
+      "step": 32740
+    },
+    {
+      "epoch": 98.35,
+      "grad_norm": 10.887227058410645,
+      "learning_rate": 6.721721721721722e-06,
+      "loss": 0.4321,
+      "step": 32750
+    },
+    {
+      "epoch": 98.38,
+      "grad_norm": 15.4978666305542,
+      "learning_rate": 6.720720720720721e-06,
+      "loss": 0.515,
+      "step": 32760
+    },
+    {
+      "epoch": 98.41,
+      "grad_norm": 14.562806129455566,
+      "learning_rate": 6.719719719719721e-06,
+      "loss": 0.4696,
+      "step": 32770
+    },
+    {
+      "epoch": 98.44,
+      "grad_norm": 27.746004104614258,
+      "learning_rate": 6.718718718718719e-06,
+      "loss": 0.4499,
+      "step": 32780
+    },
+    {
+      "epoch": 98.47,
+      "grad_norm": 13.533453941345215,
+      "learning_rate": 6.717717717717718e-06,
+      "loss": 0.496,
+      "step": 32790
+    },
+    {
+      "epoch": 98.5,
+      "grad_norm": 15.474034309387207,
+      "learning_rate": 6.716716716716717e-06,
+      "loss": 0.492,
+      "step": 32800
+    },
+    {
+      "epoch": 98.53,
+      "grad_norm": 14.723959922790527,
+      "learning_rate": 6.715715715715716e-06,
+      "loss": 0.4669,
+      "step": 32810
+    },
+    {
+      "epoch": 98.56,
+      "grad_norm": 12.52029037475586,
+      "learning_rate": 6.7147147147147155e-06,
+      "loss": 0.4278,
+      "step": 32820
+    },
+    {
+      "epoch": 98.59,
+      "grad_norm": 12.730308532714844,
+      "learning_rate": 6.713713713713714e-06,
+      "loss": 0.4535,
+      "step": 32830
+    },
+    {
+      "epoch": 98.62,
+      "grad_norm": 17.942567825317383,
+      "learning_rate": 6.7127127127127135e-06,
+      "loss": 0.4735,
+      "step": 32840
+    },
+    {
+      "epoch": 98.65,
+      "grad_norm": 12.078842163085938,
+      "learning_rate": 6.711711711711713e-06,
+      "loss": 0.4325,
+      "step": 32850
+    },
+    {
+      "epoch": 98.68,
+      "grad_norm": 13.005814552307129,
+      "learning_rate": 6.7107107107107115e-06,
+      "loss": 0.5181,
+      "step": 32860
+    },
+    {
+      "epoch": 98.71,
+      "grad_norm": 14.495569229125977,
+      "learning_rate": 6.709709709709711e-06,
+      "loss": 0.5012,
+      "step": 32870
+    },
+    {
+      "epoch": 98.74,
+      "grad_norm": 12.721776008605957,
+      "learning_rate": 6.708708708708709e-06,
+      "loss": 0.4888,
+      "step": 32880
+    },
+    {
+      "epoch": 98.77,
+      "grad_norm": 17.317598342895508,
+      "learning_rate": 6.707707707707708e-06,
+      "loss": 0.4908,
+      "step": 32890
+    },
+    {
+      "epoch": 98.8,
+      "grad_norm": 18.921810150146484,
+      "learning_rate": 6.706706706706707e-06,
+      "loss": 0.4945,
+      "step": 32900
+    },
+    {
+      "epoch": 98.83,
+      "grad_norm": 13.308624267578125,
+      "learning_rate": 6.705705705705706e-06,
+      "loss": 0.4733,
+      "step": 32910
+    },
+    {
+      "epoch": 98.86,
+      "grad_norm": 15.948408126831055,
+      "learning_rate": 6.704704704704706e-06,
+      "loss": 0.5172,
+      "step": 32920
+    },
+    {
+      "epoch": 98.89,
+      "grad_norm": 16.975757598876953,
+      "learning_rate": 6.703703703703704e-06,
+      "loss": 0.485,
+      "step": 32930
+    },
+    {
+      "epoch": 98.92,
+      "grad_norm": 16.745975494384766,
+      "learning_rate": 6.702702702702704e-06,
+      "loss": 0.5048,
+      "step": 32940
+    },
+    {
+      "epoch": 98.95,
+      "grad_norm": 18.185665130615234,
+      "learning_rate": 6.701701701701703e-06,
+      "loss": 0.4651,
+      "step": 32950
+    },
+    {
+      "epoch": 98.98,
+      "grad_norm": 15.636577606201172,
+      "learning_rate": 6.700700700700701e-06,
+      "loss": 0.49,
+      "step": 32960
+    },
+    {
+      "epoch": 99.0,
+      "eval_accuracy": 0.8663,
+      "eval_loss": 0.48792245984077454,
+      "eval_runtime": 12.6828,
+      "eval_samples_per_second": 788.467,
+      "eval_steps_per_second": 3.154,
+      "step": 32967
+    },
+    {
+      "epoch": 99.01,
+      "grad_norm": 16.06671142578125,
+      "learning_rate": 6.6996996996997e-06,
+      "loss": 0.512,
+      "step": 32970
+    },
+    {
+      "epoch": 99.04,
+      "grad_norm": 14.846404075622559,
+      "learning_rate": 6.698698698698699e-06,
+      "loss": 0.4952,
+      "step": 32980
+    },
+    {
+      "epoch": 99.07,
+      "grad_norm": 12.39246940612793,
+      "learning_rate": 6.697697697697698e-06,
+      "loss": 0.4471,
+      "step": 32990
+    },
+    {
+      "epoch": 99.1,
+      "grad_norm": 17.162826538085938,
+      "learning_rate": 6.696696696696697e-06,
+      "loss": 0.3894,
+      "step": 33000
+    },
+    {
+      "epoch": 99.13,
+      "grad_norm": 14.60906982421875,
+      "learning_rate": 6.695695695695696e-06,
+      "loss": 0.4756,
+      "step": 33010
+    },
+    {
+      "epoch": 99.16,
+      "grad_norm": 17.254634857177734,
+      "learning_rate": 6.694694694694696e-06,
+      "loss": 0.4598,
+      "step": 33020
+    },
+    {
+      "epoch": 99.19,
+      "grad_norm": 17.42293357849121,
+      "learning_rate": 6.693693693693694e-06,
+      "loss": 0.4628,
+      "step": 33030
+    },
+    {
+      "epoch": 99.22,
+      "grad_norm": 13.674723625183105,
+      "learning_rate": 6.692692692692693e-06,
+      "loss": 0.477,
+      "step": 33040
+    },
+    {
+      "epoch": 99.25,
+      "grad_norm": 12.252667427062988,
+      "learning_rate": 6.691691691691692e-06,
+      "loss": 0.4481,
+      "step": 33050
+    },
+    {
+      "epoch": 99.28,
+      "grad_norm": 14.244458198547363,
+      "learning_rate": 6.690690690690691e-06,
+      "loss": 0.5067,
+      "step": 33060
+    },
+    {
+      "epoch": 99.31,
+      "grad_norm": 17.73563003540039,
+      "learning_rate": 6.6896896896896905e-06,
+      "loss": 0.4771,
+      "step": 33070
+    },
+    {
+      "epoch": 99.34,
+      "grad_norm": 18.288068771362305,
+      "learning_rate": 6.688688688688689e-06,
+      "loss": 0.4414,
+      "step": 33080
+    },
+    {
+      "epoch": 99.37,
+      "grad_norm": 15.589969635009766,
+      "learning_rate": 6.6876876876876885e-06,
+      "loss": 0.5249,
+      "step": 33090
+    },
+    {
+      "epoch": 99.4,
+      "grad_norm": 25.752347946166992,
+      "learning_rate": 6.686686686686687e-06,
+      "loss": 0.4835,
+      "step": 33100
+    },
+    {
+      "epoch": 99.43,
+      "grad_norm": 16.232980728149414,
+      "learning_rate": 6.6856856856856866e-06,
+      "loss": 0.3953,
+      "step": 33110
+    },
+    {
+      "epoch": 99.46,
+      "grad_norm": 14.440961837768555,
+      "learning_rate": 6.684684684684684e-06,
+      "loss": 0.4616,
+      "step": 33120
+    },
+    {
+      "epoch": 99.49,
+      "grad_norm": 13.01412582397461,
+      "learning_rate": 6.683683683683684e-06,
+      "loss": 0.4697,
+      "step": 33130
+    },
+    {
+      "epoch": 99.52,
+      "grad_norm": 14.383147239685059,
+      "learning_rate": 6.682682682682683e-06,
+      "loss": 0.4731,
+      "step": 33140
+    },
+    {
+      "epoch": 99.55,
+      "grad_norm": 14.084709167480469,
+      "learning_rate": 6.681681681681682e-06,
+      "loss": 0.4689,
+      "step": 33150
+    },
+    {
+      "epoch": 99.58,
+      "grad_norm": 14.316886901855469,
+      "learning_rate": 6.680680680680681e-06,
+      "loss": 0.4921,
+      "step": 33160
+    },
+    {
+      "epoch": 99.61,
+      "grad_norm": 16.90093231201172,
+      "learning_rate": 6.679679679679681e-06,
+      "loss": 0.477,
+      "step": 33170
+    },
+    {
+      "epoch": 99.64,
+      "grad_norm": 17.076799392700195,
+      "learning_rate": 6.678678678678679e-06,
+      "loss": 0.4769,
+      "step": 33180
+    },
+    {
+      "epoch": 99.67,
+      "grad_norm": 17.553043365478516,
+      "learning_rate": 6.677677677677679e-06,
+      "loss": 0.4436,
+      "step": 33190
+    },
+    {
+      "epoch": 99.7,
+      "grad_norm": 15.456626892089844,
+      "learning_rate": 6.676676676676678e-06,
+      "loss": 0.4742,
+      "step": 33200
+    },
+    {
+      "epoch": 99.73,
+      "grad_norm": 9.593998908996582,
+      "learning_rate": 6.675675675675676e-06,
+      "loss": 0.4284,
+      "step": 33210
+    },
+    {
+      "epoch": 99.76,
+      "grad_norm": 15.17247200012207,
+      "learning_rate": 6.6746746746746745e-06,
+      "loss": 0.4006,
+      "step": 33220
+    },
+    {
+      "epoch": 99.79,
+      "grad_norm": 19.60617446899414,
+      "learning_rate": 6.673673673673674e-06,
+      "loss": 0.5058,
+      "step": 33230
+    },
+    {
+      "epoch": 99.82,
+      "grad_norm": 12.000076293945312,
+      "learning_rate": 6.672672672672673e-06,
+      "loss": 0.4474,
+      "step": 33240
+    },
+    {
+      "epoch": 99.85,
+      "grad_norm": 18.868730545043945,
+      "learning_rate": 6.671671671671672e-06,
+      "loss": 0.4497,
+      "step": 33250
+    },
+    {
+      "epoch": 99.88,
+      "grad_norm": 13.896286964416504,
+      "learning_rate": 6.670670670670671e-06,
+      "loss": 0.4907,
+      "step": 33260
+    },
+    {
+      "epoch": 99.91,
+      "grad_norm": 11.592584609985352,
+      "learning_rate": 6.669669669669671e-06,
+      "loss": 0.4269,
+      "step": 33270
+    },
+    {
+      "epoch": 99.94,
+      "grad_norm": 12.050934791564941,
+      "learning_rate": 6.6686686686686695e-06,
+      "loss": 0.459,
+      "step": 33280
+    },
+    {
+      "epoch": 99.97,
+      "grad_norm": 13.525059700012207,
+      "learning_rate": 6.667667667667668e-06,
+      "loss": 0.4687,
+      "step": 33290
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 121.43165588378906,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4886,
+      "step": 33300
+    },
+    {
+      "epoch": 100.0,
+      "eval_accuracy": 0.869,
+      "eval_loss": 0.48589110374450684,
+      "eval_runtime": 12.8203,
+      "eval_samples_per_second": 780.014,
+      "eval_steps_per_second": 3.12,
+      "step": 33300
+    },
+    {
+      "epoch": 100.03,
+      "grad_norm": 16.907848358154297,
+      "learning_rate": 6.665665665665666e-06,
+      "loss": 0.4225,
+      "step": 33310
+    },
+    {
+      "epoch": 100.06,
+      "grad_norm": 12.48476505279541,
+      "learning_rate": 6.664664664664665e-06,
+      "loss": 0.4488,
+      "step": 33320
+    },
+    {
+      "epoch": 100.09,
+      "grad_norm": 15.217780113220215,
+      "learning_rate": 6.663663663663664e-06,
+      "loss": 0.4759,
+      "step": 33330
+    },
+    {
+      "epoch": 100.12,
+      "grad_norm": 13.629714965820312,
+      "learning_rate": 6.6626626626626636e-06,
+      "loss": 0.4685,
+      "step": 33340
+    },
+    {
+      "epoch": 100.15,
+      "grad_norm": 18.99867820739746,
+      "learning_rate": 6.661661661661662e-06,
+      "loss": 0.4583,
+      "step": 33350
+    },
+    {
+      "epoch": 100.18,
+      "grad_norm": 11.75751781463623,
+      "learning_rate": 6.660660660660662e-06,
+      "loss": 0.4571,
+      "step": 33360
+    },
+    {
+      "epoch": 100.21,
+      "grad_norm": 13.465758323669434,
+      "learning_rate": 6.659659659659659e-06,
+      "loss": 0.4611,
+      "step": 33370
+    },
+    {
+      "epoch": 100.24,
+      "grad_norm": 10.934338569641113,
+      "learning_rate": 6.658658658658659e-06,
+      "loss": 0.4748,
+      "step": 33380
+    },
+    {
+      "epoch": 100.27,
+      "grad_norm": 10.475833892822266,
+      "learning_rate": 6.657657657657658e-06,
+      "loss": 0.4483,
+      "step": 33390
+    },
+    {
+      "epoch": 100.3,
+      "grad_norm": 16.67340850830078,
+      "learning_rate": 6.656656656656657e-06,
+      "loss": 0.4137,
+      "step": 33400
+    },
+    {
+      "epoch": 100.33,
+      "grad_norm": 12.555862426757812,
+      "learning_rate": 6.655655655655656e-06,
+      "loss": 0.4605,
+      "step": 33410
+    },
+    {
+      "epoch": 100.36,
+      "grad_norm": 21.453567504882812,
+      "learning_rate": 6.654654654654656e-06,
+      "loss": 0.5083,
+      "step": 33420
+    },
+    {
+      "epoch": 100.39,
+      "grad_norm": 12.915637969970703,
+      "learning_rate": 6.653653653653654e-06,
+      "loss": 0.437,
+      "step": 33430
+    },
+    {
+      "epoch": 100.42,
+      "grad_norm": 22.564815521240234,
+      "learning_rate": 6.652652652652654e-06,
+      "loss": 0.4767,
+      "step": 33440
+    },
+    {
+      "epoch": 100.45,
+      "grad_norm": 13.45366382598877,
+      "learning_rate": 6.651651651651652e-06,
+      "loss": 0.4274,
+      "step": 33450
+    },
+    {
+      "epoch": 100.48,
+      "grad_norm": 22.167945861816406,
+      "learning_rate": 6.650650650650651e-06,
+      "loss": 0.5046,
+      "step": 33460
+    },
+    {
+      "epoch": 100.51,
+      "grad_norm": 17.202890396118164,
+      "learning_rate": 6.6496496496496495e-06,
+      "loss": 0.4553,
+      "step": 33470
+    },
+    {
+      "epoch": 100.54,
+      "grad_norm": 23.579240798950195,
+      "learning_rate": 6.648648648648649e-06,
+      "loss": 0.4558,
+      "step": 33480
+    },
+    {
+      "epoch": 100.57,
+      "grad_norm": 12.670490264892578,
+      "learning_rate": 6.647647647647648e-06,
+      "loss": 0.471,
+      "step": 33490
+    },
+    {
+      "epoch": 100.6,
+      "grad_norm": 20.069271087646484,
+      "learning_rate": 6.646646646646647e-06,
+      "loss": 0.4993,
+      "step": 33500
+    },
+    {
+      "epoch": 100.63,
+      "grad_norm": 21.995807647705078,
+      "learning_rate": 6.6456456456456465e-06,
+      "loss": 0.4669,
+      "step": 33510
+    },
+    {
+      "epoch": 100.66,
+      "grad_norm": 9.638178825378418,
+      "learning_rate": 6.644644644644646e-06,
+      "loss": 0.4324,
+      "step": 33520
+    },
+    {
+      "epoch": 100.69,
+      "grad_norm": 16.4031982421875,
+      "learning_rate": 6.6436436436436445e-06,
+      "loss": 0.4879,
+      "step": 33530
+    },
+    {
+      "epoch": 100.72,
+      "grad_norm": 22.556039810180664,
+      "learning_rate": 6.642642642642643e-06,
+      "loss": 0.4921,
+      "step": 33540
+    },
+    {
+      "epoch": 100.75,
+      "grad_norm": 16.580738067626953,
+      "learning_rate": 6.641641641641642e-06,
+      "loss": 0.446,
+      "step": 33550
+    },
+    {
+      "epoch": 100.78,
+      "grad_norm": 12.405972480773926,
+      "learning_rate": 6.640640640640641e-06,
+      "loss": 0.4549,
+      "step": 33560
+    },
+    {
+      "epoch": 100.81,
+      "grad_norm": 15.59947395324707,
+      "learning_rate": 6.63963963963964e-06,
+      "loss": 0.4327,
+      "step": 33570
+    },
+    {
+      "epoch": 100.84,
+      "grad_norm": 15.44453239440918,
+      "learning_rate": 6.638638638638639e-06,
+      "loss": 0.4631,
+      "step": 33580
+    },
+    {
+      "epoch": 100.87,
+      "grad_norm": 27.959558486938477,
+      "learning_rate": 6.637637637637639e-06,
+      "loss": 0.4376,
+      "step": 33590
+    },
+    {
+      "epoch": 100.9,
+      "grad_norm": 17.501691818237305,
+      "learning_rate": 6.636636636636637e-06,
+      "loss": 0.5024,
+      "step": 33600
+    },
+    {
+      "epoch": 100.93,
+      "grad_norm": 11.904374122619629,
+      "learning_rate": 6.635635635635637e-06,
+      "loss": 0.4944,
+      "step": 33610
+    },
+    {
+      "epoch": 100.96,
+      "grad_norm": 16.482421875,
+      "learning_rate": 6.634634634634634e-06,
+      "loss": 0.4449,
+      "step": 33620
+    },
+    {
+      "epoch": 100.99,
+      "grad_norm": 12.773212432861328,
+      "learning_rate": 6.633633633633634e-06,
+      "loss": 0.4763,
+      "step": 33630
+    },
+    {
+      "epoch": 101.0,
+      "eval_accuracy": 0.868,
+      "eval_loss": 0.48400452733039856,
+      "eval_runtime": 12.7439,
+      "eval_samples_per_second": 784.691,
+      "eval_steps_per_second": 3.139,
+      "step": 33633
+    },
+    {
+      "epoch": 101.02,
+      "grad_norm": 11.47655963897705,
+      "learning_rate": 6.632632632632633e-06,
+      "loss": 0.6333,
+      "step": 33640
+    },
+    {
+      "epoch": 101.05,
+      "grad_norm": 10.327366828918457,
+      "learning_rate": 6.631631631631632e-06,
+      "loss": 0.4927,
+      "step": 33650
+    },
+    {
+      "epoch": 101.08,
+      "grad_norm": 15.850168228149414,
+      "learning_rate": 6.630630630630631e-06,
+      "loss": 0.4772,
+      "step": 33660
+    },
+    {
+      "epoch": 101.11,
+      "grad_norm": 13.528112411499023,
+      "learning_rate": 6.62962962962963e-06,
+      "loss": 0.457,
+      "step": 33670
+    },
+    {
+      "epoch": 101.14,
+      "grad_norm": 14.07089900970459,
+      "learning_rate": 6.628628628628629e-06,
+      "loss": 0.4743,
+      "step": 33680
+    },
+    {
+      "epoch": 101.17,
+      "grad_norm": 16.263784408569336,
+      "learning_rate": 6.627627627627629e-06,
+      "loss": 0.49,
+      "step": 33690
+    },
+    {
+      "epoch": 101.2,
+      "grad_norm": 20.73236656188965,
+      "learning_rate": 6.626626626626627e-06,
+      "loss": 0.4742,
+      "step": 33700
+    },
+    {
+      "epoch": 101.23,
+      "grad_norm": 21.702777862548828,
+      "learning_rate": 6.625625625625626e-06,
+      "loss": 0.4503,
+      "step": 33710
+    },
+    {
+      "epoch": 101.26,
+      "grad_norm": 20.954265594482422,
+      "learning_rate": 6.6246246246246246e-06,
+      "loss": 0.457,
+      "step": 33720
+    },
+    {
+      "epoch": 101.29,
+      "grad_norm": 10.18935775756836,
+      "learning_rate": 6.623623623623624e-06,
+      "loss": 0.4462,
+      "step": 33730
+    },
+    {
+      "epoch": 101.32,
+      "grad_norm": 15.769709587097168,
+      "learning_rate": 6.6226226226226235e-06,
+      "loss": 0.4719,
+      "step": 33740
+    },
+    {
+      "epoch": 101.35,
+      "grad_norm": 14.047489166259766,
+      "learning_rate": 6.621621621621622e-06,
+      "loss": 0.4645,
+      "step": 33750
+    },
+    {
+      "epoch": 101.38,
+      "grad_norm": 11.327953338623047,
+      "learning_rate": 6.6206206206206215e-06,
+      "loss": 0.4075,
+      "step": 33760
+    },
+    {
+      "epoch": 101.41,
+      "grad_norm": 15.141202926635742,
+      "learning_rate": 6.619619619619621e-06,
+      "loss": 0.428,
+      "step": 33770
+    },
+    {
+      "epoch": 101.44,
+      "grad_norm": 19.428138732910156,
+      "learning_rate": 6.6186186186186195e-06,
+      "loss": 0.5085,
+      "step": 33780
+    },
+    {
+      "epoch": 101.47,
+      "grad_norm": 25.5225830078125,
+      "learning_rate": 6.617617617617617e-06,
+      "loss": 0.4913,
+      "step": 33790
+    },
+    {
+      "epoch": 101.5,
+      "grad_norm": 14.827417373657227,
+      "learning_rate": 6.616616616616617e-06,
+      "loss": 0.4293,
+      "step": 33800
+    },
+    {
+      "epoch": 101.53,
+      "grad_norm": 12.409762382507324,
+      "learning_rate": 6.615615615615616e-06,
+      "loss": 0.4726,
+      "step": 33810
+    },
+    {
+      "epoch": 101.56,
+      "grad_norm": 18.56845474243164,
+      "learning_rate": 6.614614614614615e-06,
+      "loss": 0.4305,
+      "step": 33820
+    },
+    {
+      "epoch": 101.59,
+      "grad_norm": 18.675416946411133,
+      "learning_rate": 6.613613613613614e-06,
+      "loss": 0.4926,
+      "step": 33830
+    },
+    {
+      "epoch": 101.62,
+      "grad_norm": 16.639816284179688,
+      "learning_rate": 6.612612612612614e-06,
+      "loss": 0.48,
+      "step": 33840
+    },
+    {
+      "epoch": 101.65,
+      "grad_norm": 22.70361328125,
+      "learning_rate": 6.611611611611612e-06,
+      "loss": 0.5024,
+      "step": 33850
+    },
+    {
+      "epoch": 101.68,
+      "grad_norm": 17.791372299194336,
+      "learning_rate": 6.610610610610612e-06,
+      "loss": 0.4356,
+      "step": 33860
+    },
+    {
+      "epoch": 101.71,
+      "grad_norm": 25.313167572021484,
+      "learning_rate": 6.609609609609609e-06,
+      "loss": 0.4989,
+      "step": 33870
+    },
+    {
+      "epoch": 101.74,
+      "grad_norm": 13.400091171264648,
+      "learning_rate": 6.608608608608609e-06,
+      "loss": 0.4768,
+      "step": 33880
+    },
+    {
+      "epoch": 101.77,
+      "grad_norm": 15.324139595031738,
+      "learning_rate": 6.607607607607608e-06,
+      "loss": 0.4364,
+      "step": 33890
+    },
+    {
+      "epoch": 101.8,
+      "grad_norm": 20.830942153930664,
+      "learning_rate": 6.606606606606607e-06,
+      "loss": 0.461,
+      "step": 33900
+    },
+    {
+      "epoch": 101.83,
+      "grad_norm": 15.683683395385742,
+      "learning_rate": 6.605605605605606e-06,
+      "loss": 0.442,
+      "step": 33910
+    },
+    {
+      "epoch": 101.86,
+      "grad_norm": 21.630590438842773,
+      "learning_rate": 6.604604604604605e-06,
+      "loss": 0.4241,
+      "step": 33920
+    },
+    {
+      "epoch": 101.89,
+      "grad_norm": 11.24582290649414,
+      "learning_rate": 6.603603603603604e-06,
+      "loss": 0.4624,
+      "step": 33930
+    },
+    {
+      "epoch": 101.92,
+      "grad_norm": 17.067501068115234,
+      "learning_rate": 6.602602602602604e-06,
+      "loss": 0.4263,
+      "step": 33940
+    },
+    {
+      "epoch": 101.95,
+      "grad_norm": 14.197015762329102,
+      "learning_rate": 6.601601601601602e-06,
+      "loss": 0.423,
+      "step": 33950
+    },
+    {
+      "epoch": 101.98,
+      "grad_norm": 12.756484031677246,
+      "learning_rate": 6.600600600600601e-06,
+      "loss": 0.5143,
+      "step": 33960
+    },
+    {
+      "epoch": 102.0,
+      "eval_accuracy": 0.8673,
+      "eval_loss": 0.49403613805770874,
+      "eval_runtime": 12.8896,
+      "eval_samples_per_second": 775.816,
+      "eval_steps_per_second": 3.103,
+      "step": 33966
+    },
+    {
+      "epoch": 102.01,
+      "grad_norm": 14.004434585571289,
+      "learning_rate": 6.5995995995996e-06,
+      "loss": 0.4622,
+      "step": 33970
+    },
+    {
+      "epoch": 102.04,
+      "grad_norm": 21.447397232055664,
+      "learning_rate": 6.598598598598599e-06,
+      "loss": 0.4452,
+      "step": 33980
+    },
+    {
+      "epoch": 102.07,
+      "grad_norm": 12.126116752624512,
+      "learning_rate": 6.5975975975975985e-06,
+      "loss": 0.4427,
+      "step": 33990
+    },
+    {
+      "epoch": 102.1,
+      "grad_norm": 13.633326530456543,
+      "learning_rate": 6.596596596596597e-06,
+      "loss": 0.435,
+      "step": 34000
+    },
+    {
+      "epoch": 102.13,
+      "grad_norm": 13.774630546569824,
+      "learning_rate": 6.5955955955955965e-06,
+      "loss": 0.4445,
+      "step": 34010
+    },
+    {
+      "epoch": 102.16,
+      "grad_norm": 20.862611770629883,
+      "learning_rate": 6.594594594594595e-06,
+      "loss": 0.4451,
+      "step": 34020
+    },
+    {
+      "epoch": 102.19,
+      "grad_norm": 11.038888931274414,
+      "learning_rate": 6.5935935935935946e-06,
+      "loss": 0.4984,
+      "step": 34030
+    },
+    {
+      "epoch": 102.22,
+      "grad_norm": 14.033559799194336,
+      "learning_rate": 6.592592592592592e-06,
+      "loss": 0.4705,
+      "step": 34040
+    },
+    {
+      "epoch": 102.25,
+      "grad_norm": 18.631216049194336,
+      "learning_rate": 6.591591591591592e-06,
+      "loss": 0.4103,
+      "step": 34050
+    },
+    {
+      "epoch": 102.28,
+      "grad_norm": 11.53691291809082,
+      "learning_rate": 6.590590590590591e-06,
+      "loss": 0.5223,
+      "step": 34060
+    },
+    {
+      "epoch": 102.31,
+      "grad_norm": 14.756192207336426,
+      "learning_rate": 6.58958958958959e-06,
+      "loss": 0.4783,
+      "step": 34070
+    },
+    {
+      "epoch": 102.34,
+      "grad_norm": 20.108610153198242,
+      "learning_rate": 6.588588588588589e-06,
+      "loss": 0.4401,
+      "step": 34080
+    },
+    {
+      "epoch": 102.37,
+      "grad_norm": 21.488941192626953,
+      "learning_rate": 6.587587587587589e-06,
+      "loss": 0.4861,
+      "step": 34090
+    },
+    {
+      "epoch": 102.4,
+      "grad_norm": 19.676244735717773,
+      "learning_rate": 6.586586586586587e-06,
+      "loss": 0.4567,
+      "step": 34100
+    },
+    {
+      "epoch": 102.43,
+      "grad_norm": 13.662117958068848,
+      "learning_rate": 6.585585585585587e-06,
+      "loss": 0.4425,
+      "step": 34110
+    },
+    {
+      "epoch": 102.46,
+      "grad_norm": 23.70462989807129,
+      "learning_rate": 6.5845845845845845e-06,
+      "loss": 0.4804,
+      "step": 34120
+    },
+    {
+      "epoch": 102.49,
+      "grad_norm": 21.31736946105957,
+      "learning_rate": 6.583583583583584e-06,
+      "loss": 0.4648,
+      "step": 34130
+    },
+    {
+      "epoch": 102.52,
+      "grad_norm": 15.793845176696777,
+      "learning_rate": 6.5825825825825825e-06,
+      "loss": 0.4864,
+      "step": 34140
+    },
+    {
+      "epoch": 102.55,
+      "grad_norm": 15.998614311218262,
+      "learning_rate": 6.581581581581582e-06,
+      "loss": 0.4458,
+      "step": 34150
+    },
+    {
+      "epoch": 102.58,
+      "grad_norm": 16.32208251953125,
+      "learning_rate": 6.580580580580581e-06,
+      "loss": 0.4666,
+      "step": 34160
+    },
+    {
+      "epoch": 102.61,
+      "grad_norm": 15.102506637573242,
+      "learning_rate": 6.57957957957958e-06,
+      "loss": 0.4363,
+      "step": 34170
+    },
+    {
+      "epoch": 102.64,
+      "grad_norm": 15.43324089050293,
+      "learning_rate": 6.578578578578579e-06,
+      "loss": 0.4757,
+      "step": 34180
+    },
+    {
+      "epoch": 102.67,
+      "grad_norm": 15.57242202758789,
+      "learning_rate": 6.577577577577579e-06,
+      "loss": 0.4506,
+      "step": 34190
+    },
+    {
+      "epoch": 102.7,
+      "grad_norm": 18.14263343811035,
+      "learning_rate": 6.5765765765765775e-06,
+      "loss": 0.4464,
+      "step": 34200
+    },
+    {
+      "epoch": 102.73,
+      "grad_norm": 18.687976837158203,
+      "learning_rate": 6.575575575575576e-06,
+      "loss": 0.4288,
+      "step": 34210
+    },
+    {
+      "epoch": 102.76,
+      "grad_norm": 14.506706237792969,
+      "learning_rate": 6.574574574574575e-06,
+      "loss": 0.4357,
+      "step": 34220
+    },
+    {
+      "epoch": 102.79,
+      "grad_norm": 13.61119556427002,
+      "learning_rate": 6.573573573573574e-06,
+      "loss": 0.419,
+      "step": 34230
+    },
+    {
+      "epoch": 102.82,
+      "grad_norm": 12.875226020812988,
+      "learning_rate": 6.572572572572573e-06,
+      "loss": 0.4554,
+      "step": 34240
+    },
+    {
+      "epoch": 102.85,
+      "grad_norm": 18.31559181213379,
+      "learning_rate": 6.571571571571572e-06,
+      "loss": 0.4879,
+      "step": 34250
+    },
+    {
+      "epoch": 102.88,
+      "grad_norm": 16.864055633544922,
+      "learning_rate": 6.5705705705705716e-06,
+      "loss": 0.4552,
+      "step": 34260
+    },
+    {
+      "epoch": 102.91,
+      "grad_norm": 11.102497100830078,
+      "learning_rate": 6.56956956956957e-06,
+      "loss": 0.5065,
+      "step": 34270
+    },
+    {
+      "epoch": 102.94,
+      "grad_norm": 18.057767868041992,
+      "learning_rate": 6.56856856856857e-06,
+      "loss": 0.5048,
+      "step": 34280
+    },
+    {
+      "epoch": 102.97,
+      "grad_norm": 11.266833305358887,
+      "learning_rate": 6.567567567567567e-06,
+      "loss": 0.4732,
+      "step": 34290
+    },
+    {
+      "epoch": 103.0,
+      "eval_accuracy": 0.8699,
+      "eval_loss": 0.48267313838005066,
+      "eval_runtime": 12.7101,
+      "eval_samples_per_second": 786.775,
+      "eval_steps_per_second": 3.147,
+      "step": 34299
+    },
+    {
+      "epoch": 103.0,
+      "grad_norm": 16.602678298950195,
+      "learning_rate": 6.566566566566567e-06,
+      "loss": 0.4724,
+      "step": 34300
+    },
+    {
+      "epoch": 103.03,
+      "grad_norm": 21.219890594482422,
+      "learning_rate": 6.565565565565566e-06,
+      "loss": 0.4368,
+      "step": 34310
+    },
+    {
+      "epoch": 103.06,
+      "grad_norm": 25.041221618652344,
+      "learning_rate": 6.564564564564565e-06,
+      "loss": 0.4541,
+      "step": 34320
+    },
+    {
+      "epoch": 103.09,
+      "grad_norm": 11.220368385314941,
+      "learning_rate": 6.563563563563564e-06,
+      "loss": 0.4922,
+      "step": 34330
+    },
+    {
+      "epoch": 103.12,
+      "grad_norm": 15.785404205322266,
+      "learning_rate": 6.562562562562564e-06,
+      "loss": 0.4979,
+      "step": 34340
+    },
+    {
+      "epoch": 103.15,
+      "grad_norm": 21.085695266723633,
+      "learning_rate": 6.561561561561562e-06,
+      "loss": 0.5156,
+      "step": 34350
+    },
+    {
+      "epoch": 103.18,
+      "grad_norm": 10.323051452636719,
+      "learning_rate": 6.560560560560562e-06,
+      "loss": 0.4742,
+      "step": 34360
+    },
+    {
+      "epoch": 103.21,
+      "grad_norm": 18.191802978515625,
+      "learning_rate": 6.5595595595595595e-06,
+      "loss": 0.4982,
+      "step": 34370
+    },
+    {
+      "epoch": 103.24,
+      "grad_norm": 14.714456558227539,
+      "learning_rate": 6.558558558558559e-06,
+      "loss": 0.4804,
+      "step": 34380
+    },
+    {
+      "epoch": 103.27,
+      "grad_norm": 17.00783348083496,
+      "learning_rate": 6.5575575575575575e-06,
+      "loss": 0.4312,
+      "step": 34390
+    },
+    {
+      "epoch": 103.3,
+      "grad_norm": 17.183042526245117,
+      "learning_rate": 6.556556556556557e-06,
+      "loss": 0.4298,
+      "step": 34400
+    },
+    {
+      "epoch": 103.33,
+      "grad_norm": 11.022048950195312,
+      "learning_rate": 6.555555555555556e-06,
+      "loss": 0.4824,
+      "step": 34410
+    },
+    {
+      "epoch": 103.36,
+      "grad_norm": 15.616822242736816,
+      "learning_rate": 6.554554554554555e-06,
+      "loss": 0.4757,
+      "step": 34420
+    },
+    {
+      "epoch": 103.39,
+      "grad_norm": 10.010004997253418,
+      "learning_rate": 6.5535535535535544e-06,
+      "loss": 0.4502,
+      "step": 34430
+    },
+    {
+      "epoch": 103.42,
+      "grad_norm": 13.172829627990723,
+      "learning_rate": 6.552552552552554e-06,
+      "loss": 0.4886,
+      "step": 34440
+    },
+    {
+      "epoch": 103.45,
+      "grad_norm": 23.104387283325195,
+      "learning_rate": 6.551551551551552e-06,
+      "loss": 0.4419,
+      "step": 34450
+    },
+    {
+      "epoch": 103.48,
+      "grad_norm": 13.12995433807373,
+      "learning_rate": 6.550550550550551e-06,
+      "loss": 0.4522,
+      "step": 34460
+    },
+    {
+      "epoch": 103.51,
+      "grad_norm": 17.28179359436035,
+      "learning_rate": 6.54954954954955e-06,
+      "loss": 0.4537,
+      "step": 34470
+    },
+    {
+      "epoch": 103.54,
+      "grad_norm": 10.82555866241455,
+      "learning_rate": 6.548548548548549e-06,
+      "loss": 0.492,
+      "step": 34480
+    },
+    {
+      "epoch": 103.57,
+      "grad_norm": 14.938183784484863,
+      "learning_rate": 6.547547547547548e-06,
+      "loss": 0.4883,
+      "step": 34490
+    },
+    {
+      "epoch": 103.6,
+      "grad_norm": 17.44671630859375,
+      "learning_rate": 6.546546546546547e-06,
+      "loss": 0.4734,
+      "step": 34500
+    },
+    {
+      "epoch": 103.63,
+      "grad_norm": 15.17357349395752,
+      "learning_rate": 6.545545545545547e-06,
+      "loss": 0.507,
+      "step": 34510
+    },
+    {
+      "epoch": 103.66,
+      "grad_norm": 20.51030731201172,
+      "learning_rate": 6.544544544544545e-06,
+      "loss": 0.5349,
+      "step": 34520
+    },
+    {
+      "epoch": 103.69,
+      "grad_norm": 20.062606811523438,
+      "learning_rate": 6.543543543543545e-06,
+      "loss": 0.4389,
+      "step": 34530
+    },
+    {
+      "epoch": 103.72,
+      "grad_norm": 11.506123542785645,
+      "learning_rate": 6.542542542542542e-06,
+      "loss": 0.462,
+      "step": 34540
+    },
+    {
+      "epoch": 103.75,
+      "grad_norm": 15.605498313903809,
+      "learning_rate": 6.541541541541542e-06,
+      "loss": 0.4388,
+      "step": 34550
+    },
+    {
+      "epoch": 103.78,
+      "grad_norm": 16.985876083374023,
+      "learning_rate": 6.540540540540541e-06,
+      "loss": 0.4693,
+      "step": 34560
+    },
+    {
+      "epoch": 103.81,
+      "grad_norm": 21.2596492767334,
+      "learning_rate": 6.53953953953954e-06,
+      "loss": 0.5101,
+      "step": 34570
+    },
+    {
+      "epoch": 103.84,
+      "grad_norm": 17.95951271057129,
+      "learning_rate": 6.538538538538539e-06,
+      "loss": 0.4761,
+      "step": 34580
+    },
+    {
+      "epoch": 103.87,
+      "grad_norm": 16.426918029785156,
+      "learning_rate": 6.537537537537538e-06,
+      "loss": 0.4815,
+      "step": 34590
+    },
+    {
+      "epoch": 103.9,
+      "grad_norm": 21.765098571777344,
+      "learning_rate": 6.536536536536537e-06,
+      "loss": 0.4761,
+      "step": 34600
+    },
+    {
+      "epoch": 103.93,
+      "grad_norm": 11.473360061645508,
+      "learning_rate": 6.535535535535537e-06,
+      "loss": 0.445,
+      "step": 34610
+    },
+    {
+      "epoch": 103.96,
+      "grad_norm": 13.943897247314453,
+      "learning_rate": 6.5345345345345345e-06,
+      "loss": 0.4355,
+      "step": 34620
+    },
+    {
+      "epoch": 103.99,
+      "grad_norm": 13.445980072021484,
+      "learning_rate": 6.533533533533534e-06,
+      "loss": 0.481,
+      "step": 34630
+    },
+    {
+      "epoch": 104.0,
+      "eval_accuracy": 0.8686,
+      "eval_loss": 0.48906221985816956,
+      "eval_runtime": 13.6789,
+      "eval_samples_per_second": 731.053,
+      "eval_steps_per_second": 2.924,
+      "step": 34632
+    },
+    {
+      "epoch": 104.02,
+      "grad_norm": 18.762832641601562,
+      "learning_rate": 6.5325325325325326e-06,
+      "loss": 0.4947,
+      "step": 34640
+    },
+    {
+      "epoch": 104.05,
+      "grad_norm": 22.056650161743164,
+      "learning_rate": 6.531531531531532e-06,
+      "loss": 0.4465,
+      "step": 34650
+    },
+    {
+      "epoch": 104.08,
+      "grad_norm": 15.976855278015137,
+      "learning_rate": 6.5305305305305314e-06,
+      "loss": 0.4052,
+      "step": 34660
+    },
+    {
+      "epoch": 104.11,
+      "grad_norm": 13.380452156066895,
+      "learning_rate": 6.52952952952953e-06,
+      "loss": 0.4702,
+      "step": 34670
+    },
+    {
+      "epoch": 104.14,
+      "grad_norm": 13.828340530395508,
+      "learning_rate": 6.5285285285285295e-06,
+      "loss": 0.5217,
+      "step": 34680
+    },
+    {
+      "epoch": 104.17,
+      "grad_norm": 12.411547660827637,
+      "learning_rate": 6.527527527527528e-06,
+      "loss": 0.5126,
+      "step": 34690
+    },
+    {
+      "epoch": 104.2,
+      "grad_norm": 16.11130714416504,
+      "learning_rate": 6.526526526526527e-06,
+      "loss": 0.4808,
+      "step": 34700
+    },
+    {
+      "epoch": 104.23,
+      "grad_norm": 13.44447135925293,
+      "learning_rate": 6.525525525525525e-06,
+      "loss": 0.4152,
+      "step": 34710
+    },
+    {
+      "epoch": 104.26,
+      "grad_norm": 16.086883544921875,
+      "learning_rate": 6.524524524524525e-06,
+      "loss": 0.4703,
+      "step": 34720
+    },
+    {
+      "epoch": 104.29,
+      "grad_norm": 13.868494987487793,
+      "learning_rate": 6.523523523523524e-06,
+      "loss": 0.4556,
+      "step": 34730
+    },
+    {
+      "epoch": 104.32,
+      "grad_norm": 16.33922004699707,
+      "learning_rate": 6.522522522522523e-06,
+      "loss": 0.4051,
+      "step": 34740
+    },
+    {
+      "epoch": 104.35,
+      "grad_norm": 18.28533363342285,
+      "learning_rate": 6.521521521521522e-06,
+      "loss": 0.487,
+      "step": 34750
+    },
+    {
+      "epoch": 104.38,
+      "grad_norm": 17.57025718688965,
+      "learning_rate": 6.520520520520522e-06,
+      "loss": 0.4906,
+      "step": 34760
+    },
+    {
+      "epoch": 104.41,
+      "grad_norm": 21.027692794799805,
+      "learning_rate": 6.51951951951952e-06,
+      "loss": 0.4566,
+      "step": 34770
+    },
+    {
+      "epoch": 104.44,
+      "grad_norm": 15.233656883239746,
+      "learning_rate": 6.51851851851852e-06,
+      "loss": 0.5165,
+      "step": 34780
+    },
+    {
+      "epoch": 104.47,
+      "grad_norm": 12.021224021911621,
+      "learning_rate": 6.517517517517517e-06,
+      "loss": 0.4592,
+      "step": 34790
+    },
+    {
+      "epoch": 104.5,
+      "grad_norm": 25.758718490600586,
+      "learning_rate": 6.516516516516517e-06,
+      "loss": 0.4501,
+      "step": 34800
+    },
+    {
+      "epoch": 104.53,
+      "grad_norm": 18.37145233154297,
+      "learning_rate": 6.5155155155155155e-06,
+      "loss": 0.4429,
+      "step": 34810
+    },
+    {
+      "epoch": 104.56,
+      "grad_norm": 13.933323860168457,
+      "learning_rate": 6.514514514514515e-06,
+      "loss": 0.4991,
+      "step": 34820
+    },
+    {
+      "epoch": 104.59,
+      "grad_norm": 14.943764686584473,
+      "learning_rate": 6.513513513513514e-06,
+      "loss": 0.4816,
+      "step": 34830
+    },
+    {
+      "epoch": 104.62,
+      "grad_norm": 20.590131759643555,
+      "learning_rate": 6.512512512512513e-06,
+      "loss": 0.4852,
+      "step": 34840
+    },
+    {
+      "epoch": 104.65,
+      "grad_norm": 13.435240745544434,
+      "learning_rate": 6.511511511511512e-06,
+      "loss": 0.4965,
+      "step": 34850
+    },
+    {
+      "epoch": 104.68,
+      "grad_norm": 14.319186210632324,
+      "learning_rate": 6.510510510510512e-06,
+      "loss": 0.4292,
+      "step": 34860
+    },
+    {
+      "epoch": 104.71,
+      "grad_norm": 14.052082061767578,
+      "learning_rate": 6.5095095095095096e-06,
+      "loss": 0.4346,
+      "step": 34870
+    },
+    {
+      "epoch": 104.74,
+      "grad_norm": 24.241901397705078,
+      "learning_rate": 6.508508508508509e-06,
+      "loss": 0.4641,
+      "step": 34880
+    },
+    {
+      "epoch": 104.77,
+      "grad_norm": 17.13144302368164,
+      "learning_rate": 6.507507507507508e-06,
+      "loss": 0.4686,
+      "step": 34890
+    },
+    {
+      "epoch": 104.8,
+      "grad_norm": 15.892746925354004,
+      "learning_rate": 6.506506506506507e-06,
+      "loss": 0.4462,
+      "step": 34900
+    },
+    {
+      "epoch": 104.83,
+      "grad_norm": 14.129039764404297,
+      "learning_rate": 6.5055055055055065e-06,
+      "loss": 0.4453,
+      "step": 34910
+    },
+    {
+      "epoch": 104.86,
+      "grad_norm": 16.14947509765625,
+      "learning_rate": 6.504504504504505e-06,
+      "loss": 0.4131,
+      "step": 34920
+    },
+    {
+      "epoch": 104.89,
+      "grad_norm": 22.250478744506836,
+      "learning_rate": 6.5035035035035045e-06,
+      "loss": 0.4884,
+      "step": 34930
+    },
+    {
+      "epoch": 104.92,
+      "grad_norm": 15.036622047424316,
+      "learning_rate": 6.502502502502503e-06,
+      "loss": 0.4291,
+      "step": 34940
+    },
+    {
+      "epoch": 104.95,
+      "grad_norm": 14.797932624816895,
+      "learning_rate": 6.501501501501502e-06,
+      "loss": 0.4515,
+      "step": 34950
+    },
+    {
+      "epoch": 104.98,
+      "grad_norm": 19.132328033447266,
+      "learning_rate": 6.5005005005005e-06,
+      "loss": 0.5015,
+      "step": 34960
+    },
+    {
+      "epoch": 105.0,
+      "eval_accuracy": 0.8651,
+      "eval_loss": 0.5004428625106812,
+      "eval_runtime": 13.2835,
+      "eval_samples_per_second": 752.813,
+      "eval_steps_per_second": 3.011,
+      "step": 34965
+    },
+    {
+      "epoch": 105.02,
+      "grad_norm": 21.48062515258789,
+      "learning_rate": 6.4994994994995e-06,
+      "loss": 0.4042,
+      "step": 34970
+    },
+    {
+      "epoch": 105.05,
+      "grad_norm": 11.162670135498047,
+      "learning_rate": 6.498498498498499e-06,
+      "loss": 0.5015,
+      "step": 34980
+    },
+    {
+      "epoch": 105.08,
+      "grad_norm": 13.20266056060791,
+      "learning_rate": 6.497497497497498e-06,
+      "loss": 0.4062,
+      "step": 34990
+    },
+    {
+      "epoch": 105.11,
+      "grad_norm": 15.322538375854492,
+      "learning_rate": 6.496496496496497e-06,
+      "loss": 0.4287,
+      "step": 35000
+    },
+    {
+      "epoch": 105.14,
+      "grad_norm": 18.97233009338379,
+      "learning_rate": 6.495495495495497e-06,
+      "loss": 0.5014,
+      "step": 35010
+    },
+    {
+      "epoch": 105.17,
+      "grad_norm": 13.122167587280273,
+      "learning_rate": 6.494494494494495e-06,
+      "loss": 0.4069,
+      "step": 35020
+    },
+    {
+      "epoch": 105.2,
+      "grad_norm": 15.641549110412598,
+      "learning_rate": 6.493493493493495e-06,
+      "loss": 0.4165,
+      "step": 35030
+    },
+    {
+      "epoch": 105.23,
+      "grad_norm": 28.32853889465332,
+      "learning_rate": 6.4924924924924924e-06,
+      "loss": 0.4697,
+      "step": 35040
+    },
+    {
+      "epoch": 105.26,
+      "grad_norm": 13.533641815185547,
+      "learning_rate": 6.491491491491492e-06,
+      "loss": 0.4613,
+      "step": 35050
+    },
+    {
+      "epoch": 105.29,
+      "grad_norm": 12.584187507629395,
+      "learning_rate": 6.4904904904904905e-06,
+      "loss": 0.4114,
+      "step": 35060
+    },
+    {
+      "epoch": 105.32,
+      "grad_norm": 18.343921661376953,
+      "learning_rate": 6.48948948948949e-06,
+      "loss": 0.4662,
+      "step": 35070
+    },
+    {
+      "epoch": 105.35,
+      "grad_norm": 17.33719253540039,
+      "learning_rate": 6.488488488488489e-06,
+      "loss": 0.5016,
+      "step": 35080
+    },
+    {
+      "epoch": 105.38,
+      "grad_norm": 17.69260597229004,
+      "learning_rate": 6.487487487487488e-06,
+      "loss": 0.475,
+      "step": 35090
+    },
+    {
+      "epoch": 105.41,
+      "grad_norm": 11.680750846862793,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.4511,
+      "step": 35100
+    },
+    {
+      "epoch": 105.44,
+      "grad_norm": 12.34111499786377,
+      "learning_rate": 6.485485485485487e-06,
+      "loss": 0.4097,
+      "step": 35110
+    },
+    {
+      "epoch": 105.47,
+      "grad_norm": 22.025205612182617,
+      "learning_rate": 6.484484484484485e-06,
+      "loss": 0.4043,
+      "step": 35120
+    },
+    {
+      "epoch": 105.5,
+      "grad_norm": 17.74778175354004,
+      "learning_rate": 6.483483483483484e-06,
+      "loss": 0.3892,
+      "step": 35130
+    },
+    {
+      "epoch": 105.53,
+      "grad_norm": 20.969491958618164,
+      "learning_rate": 6.482482482482483e-06,
+      "loss": 0.4429,
+      "step": 35140
+    },
+    {
+      "epoch": 105.56,
+      "grad_norm": 31.45017433166504,
+      "learning_rate": 6.481481481481482e-06,
+      "loss": 0.4826,
+      "step": 35150
+    },
+    {
+      "epoch": 105.59,
+      "grad_norm": 15.831473350524902,
+      "learning_rate": 6.480480480480481e-06,
+      "loss": 0.4788,
+      "step": 35160
+    },
+    {
+      "epoch": 105.62,
+      "grad_norm": 15.13601016998291,
+      "learning_rate": 6.47947947947948e-06,
+      "loss": 0.4496,
+      "step": 35170
+    },
+    {
+      "epoch": 105.65,
+      "grad_norm": 20.015901565551758,
+      "learning_rate": 6.4784784784784796e-06,
+      "loss": 0.4705,
+      "step": 35180
+    },
+    {
+      "epoch": 105.68,
+      "grad_norm": 12.278945922851562,
+      "learning_rate": 6.477477477477478e-06,
+      "loss": 0.4633,
+      "step": 35190
+    },
+    {
+      "epoch": 105.71,
+      "grad_norm": 10.968920707702637,
+      "learning_rate": 6.476476476476477e-06,
+      "loss": 0.4687,
+      "step": 35200
+    },
+    {
+      "epoch": 105.74,
+      "grad_norm": 21.497488021850586,
+      "learning_rate": 6.475475475475475e-06,
+      "loss": 0.4057,
+      "step": 35210
+    },
+    {
+      "epoch": 105.77,
+      "grad_norm": 22.245512008666992,
+      "learning_rate": 6.474474474474475e-06,
+      "loss": 0.4921,
+      "step": 35220
+    },
+    {
+      "epoch": 105.8,
+      "grad_norm": 18.429443359375,
+      "learning_rate": 6.473473473473474e-06,
+      "loss": 0.4467,
+      "step": 35230
+    },
+    {
+      "epoch": 105.83,
+      "grad_norm": 19.91747283935547,
+      "learning_rate": 6.472472472472473e-06,
+      "loss": 0.4482,
+      "step": 35240
+    },
+    {
+      "epoch": 105.86,
+      "grad_norm": 15.55008316040039,
+      "learning_rate": 6.471471471471472e-06,
+      "loss": 0.442,
+      "step": 35250
+    },
+    {
+      "epoch": 105.89,
+      "grad_norm": 15.618000984191895,
+      "learning_rate": 6.470470470470472e-06,
+      "loss": 0.4651,
+      "step": 35260
+    },
+    {
+      "epoch": 105.92,
+      "grad_norm": 19.244646072387695,
+      "learning_rate": 6.46946946946947e-06,
+      "loss": 0.4407,
+      "step": 35270
+    },
+    {
+      "epoch": 105.95,
+      "grad_norm": 16.39017105102539,
+      "learning_rate": 6.46846846846847e-06,
+      "loss": 0.4724,
+      "step": 35280
+    },
+    {
+      "epoch": 105.98,
+      "grad_norm": 24.12604522705078,
+      "learning_rate": 6.4674674674674675e-06,
+      "loss": 0.4596,
+      "step": 35290
+    },
+    {
+      "epoch": 106.0,
+      "eval_accuracy": 0.8669,
+      "eval_loss": 0.49500948190689087,
+      "eval_runtime": 13.5306,
+      "eval_samples_per_second": 739.067,
+      "eval_steps_per_second": 2.956,
+      "step": 35298
+    },
+    {
+      "epoch": 106.01,
+      "grad_norm": 13.878896713256836,
+      "learning_rate": 6.466466466466467e-06,
+      "loss": 0.5043,
+      "step": 35300
+    },
+    {
+      "epoch": 106.04,
+      "grad_norm": 17.271507263183594,
+      "learning_rate": 6.4654654654654655e-06,
+      "loss": 0.429,
+      "step": 35310
+    },
+    {
+      "epoch": 106.07,
+      "grad_norm": 19.360624313354492,
+      "learning_rate": 6.464464464464465e-06,
+      "loss": 0.4447,
+      "step": 35320
+    },
+    {
+      "epoch": 106.1,
+      "grad_norm": 15.939460754394531,
+      "learning_rate": 6.463463463463464e-06,
+      "loss": 0.4936,
+      "step": 35330
+    },
+    {
+      "epoch": 106.13,
+      "grad_norm": 15.272868156433105,
+      "learning_rate": 6.462462462462463e-06,
+      "loss": 0.4829,
+      "step": 35340
+    },
+    {
+      "epoch": 106.16,
+      "grad_norm": 12.78689956665039,
+      "learning_rate": 6.4614614614614624e-06,
+      "loss": 0.4652,
+      "step": 35350
+    },
+    {
+      "epoch": 106.19,
+      "grad_norm": 13.677848815917969,
+      "learning_rate": 6.460460460460462e-06,
+      "loss": 0.454,
+      "step": 35360
+    },
+    {
+      "epoch": 106.22,
+      "grad_norm": 15.82308292388916,
+      "learning_rate": 6.45945945945946e-06,
+      "loss": 0.446,
+      "step": 35370
+    },
+    {
+      "epoch": 106.25,
+      "grad_norm": 18.70569610595703,
+      "learning_rate": 6.458458458458458e-06,
+      "loss": 0.4437,
+      "step": 35380
+    },
+    {
+      "epoch": 106.28,
+      "grad_norm": 17.61282730102539,
+      "learning_rate": 6.457457457457458e-06,
+      "loss": 0.4684,
+      "step": 35390
+    },
+    {
+      "epoch": 106.31,
+      "grad_norm": 11.460628509521484,
+      "learning_rate": 6.456456456456457e-06,
+      "loss": 0.4773,
+      "step": 35400
+    },
+    {
+      "epoch": 106.34,
+      "grad_norm": 19.902620315551758,
+      "learning_rate": 6.455455455455456e-06,
+      "loss": 0.4694,
+      "step": 35410
+    },
+    {
+      "epoch": 106.37,
+      "grad_norm": 11.045598983764648,
+      "learning_rate": 6.454454454454455e-06,
+      "loss": 0.5037,
+      "step": 35420
+    },
+    {
+      "epoch": 106.4,
+      "grad_norm": 13.149656295776367,
+      "learning_rate": 6.453453453453455e-06,
+      "loss": 0.5117,
+      "step": 35430
+    },
+    {
+      "epoch": 106.43,
+      "grad_norm": 15.715804100036621,
+      "learning_rate": 6.452452452452453e-06,
+      "loss": 0.4455,
+      "step": 35440
+    },
+    {
+      "epoch": 106.46,
+      "grad_norm": 12.559956550598145,
+      "learning_rate": 6.451451451451452e-06,
+      "loss": 0.4346,
+      "step": 35450
+    },
+    {
+      "epoch": 106.49,
+      "grad_norm": 24.56302261352539,
+      "learning_rate": 6.45045045045045e-06,
+      "loss": 0.4792,
+      "step": 35460
+    },
+    {
+      "epoch": 106.52,
+      "grad_norm": 23.52665901184082,
+      "learning_rate": 6.44944944944945e-06,
+      "loss": 0.4431,
+      "step": 35470
+    },
+    {
+      "epoch": 106.55,
+      "grad_norm": 16.258167266845703,
+      "learning_rate": 6.448448448448449e-06,
+      "loss": 0.4511,
+      "step": 35480
+    },
+    {
+      "epoch": 106.58,
+      "grad_norm": 14.013567924499512,
+      "learning_rate": 6.447447447447448e-06,
+      "loss": 0.4676,
+      "step": 35490
+    },
+    {
+      "epoch": 106.61,
+      "grad_norm": 23.866430282592773,
+      "learning_rate": 6.446446446446447e-06,
+      "loss": 0.4739,
+      "step": 35500
+    },
+    {
+      "epoch": 106.64,
+      "grad_norm": 20.256336212158203,
+      "learning_rate": 6.445445445445446e-06,
+      "loss": 0.4451,
+      "step": 35510
+    },
+    {
+      "epoch": 106.67,
+      "grad_norm": 13.717354774475098,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 0.4405,
+      "step": 35520
+    },
+    {
+      "epoch": 106.7,
+      "grad_norm": 13.967938423156738,
+      "learning_rate": 6.443443443443445e-06,
+      "loss": 0.46,
+      "step": 35530
+    },
+    {
+      "epoch": 106.73,
+      "grad_norm": 16.851036071777344,
+      "learning_rate": 6.4424424424424425e-06,
+      "loss": 0.4638,
+      "step": 35540
+    },
+    {
+      "epoch": 106.76,
+      "grad_norm": 14.284954071044922,
+      "learning_rate": 6.441441441441442e-06,
+      "loss": 0.5015,
+      "step": 35550
+    },
+    {
+      "epoch": 106.79,
+      "grad_norm": 24.220151901245117,
+      "learning_rate": 6.4404404404404406e-06,
+      "loss": 0.4055,
+      "step": 35560
+    },
+    {
+      "epoch": 106.82,
+      "grad_norm": 18.184396743774414,
+      "learning_rate": 6.43943943943944e-06,
+      "loss": 0.4956,
+      "step": 35570
+    },
+    {
+      "epoch": 106.85,
+      "grad_norm": 20.5683536529541,
+      "learning_rate": 6.4384384384384394e-06,
+      "loss": 0.4684,
+      "step": 35580
+    },
+    {
+      "epoch": 106.88,
+      "grad_norm": 10.556412696838379,
+      "learning_rate": 6.437437437437438e-06,
+      "loss": 0.4342,
+      "step": 35590
+    },
+    {
+      "epoch": 106.91,
+      "grad_norm": 11.02188491821289,
+      "learning_rate": 6.4364364364364375e-06,
+      "loss": 0.4107,
+      "step": 35600
+    },
+    {
+      "epoch": 106.94,
+      "grad_norm": 19.23082733154297,
+      "learning_rate": 6.435435435435436e-06,
+      "loss": 0.438,
+      "step": 35610
+    },
+    {
+      "epoch": 106.97,
+      "grad_norm": 17.896455764770508,
+      "learning_rate": 6.434434434434435e-06,
+      "loss": 0.4303,
+      "step": 35620
+    },
+    {
+      "epoch": 107.0,
+      "grad_norm": 18.535493850708008,
+      "learning_rate": 6.433433433433433e-06,
+      "loss": 0.4201,
+      "step": 35630
+    },
+    {
+      "epoch": 107.0,
+      "eval_accuracy": 0.866,
+      "eval_loss": 0.4920419454574585,
+      "eval_runtime": 12.8724,
+      "eval_samples_per_second": 776.853,
+      "eval_steps_per_second": 3.107,
+      "step": 35631
+    },
+    {
+      "epoch": 107.03,
+      "grad_norm": 16.468294143676758,
+      "learning_rate": 6.432432432432433e-06,
+      "loss": 0.4521,
+      "step": 35640
+    },
+    {
+      "epoch": 107.06,
+      "grad_norm": 16.997175216674805,
+      "learning_rate": 6.431431431431432e-06,
+      "loss": 0.4835,
+      "step": 35650
+    },
+    {
+      "epoch": 107.09,
+      "grad_norm": 19.282634735107422,
+      "learning_rate": 6.430430430430431e-06,
+      "loss": 0.4563,
+      "step": 35660
+    },
+    {
+      "epoch": 107.12,
+      "grad_norm": 18.652645111083984,
+      "learning_rate": 6.42942942942943e-06,
+      "loss": 0.4976,
+      "step": 35670
+    },
+    {
+      "epoch": 107.15,
+      "grad_norm": 14.071982383728027,
+      "learning_rate": 6.42842842842843e-06,
+      "loss": 0.4663,
+      "step": 35680
+    },
+    {
+      "epoch": 107.18,
+      "grad_norm": 16.45686912536621,
+      "learning_rate": 6.427427427427428e-06,
+      "loss": 0.4786,
+      "step": 35690
+    },
+    {
+      "epoch": 107.21,
+      "grad_norm": 12.322075843811035,
+      "learning_rate": 6.426426426426427e-06,
+      "loss": 0.4348,
+      "step": 35700
+    },
+    {
+      "epoch": 107.24,
+      "grad_norm": 11.027575492858887,
+      "learning_rate": 6.425425425425425e-06,
+      "loss": 0.4784,
+      "step": 35710
+    },
+    {
+      "epoch": 107.27,
+      "grad_norm": 22.93187713623047,
+      "learning_rate": 6.424424424424425e-06,
+      "loss": 0.4418,
+      "step": 35720
+    },
+    {
+      "epoch": 107.3,
+      "grad_norm": 16.874858856201172,
+      "learning_rate": 6.4234234234234234e-06,
+      "loss": 0.4434,
+      "step": 35730
+    },
+    {
+      "epoch": 107.33,
+      "grad_norm": 15.091069221496582,
+      "learning_rate": 6.422422422422423e-06,
+      "loss": 0.4324,
+      "step": 35740
+    },
+    {
+      "epoch": 107.36,
+      "grad_norm": 16.402076721191406,
+      "learning_rate": 6.421421421421422e-06,
+      "loss": 0.4772,
+      "step": 35750
+    },
+    {
+      "epoch": 107.39,
+      "grad_norm": 12.500945091247559,
+      "learning_rate": 6.420420420420421e-06,
+      "loss": 0.4616,
+      "step": 35760
+    },
+    {
+      "epoch": 107.42,
+      "grad_norm": 10.897164344787598,
+      "learning_rate": 6.41941941941942e-06,
+      "loss": 0.4551,
+      "step": 35770
+    },
+    {
+      "epoch": 107.45,
+      "grad_norm": 21.0148868560791,
+      "learning_rate": 6.418418418418418e-06,
+      "loss": 0.3928,
+      "step": 35780
+    },
+    {
+      "epoch": 107.48,
+      "grad_norm": 17.215126037597656,
+      "learning_rate": 6.4174174174174176e-06,
+      "loss": 0.4795,
+      "step": 35790
+    },
+    {
+      "epoch": 107.51,
+      "grad_norm": 12.059027671813965,
+      "learning_rate": 6.416416416416417e-06,
+      "loss": 0.4983,
+      "step": 35800
+    },
+    {
+      "epoch": 107.54,
+      "grad_norm": 10.076733589172363,
+      "learning_rate": 6.415415415415416e-06,
+      "loss": 0.4169,
+      "step": 35810
+    },
+    {
+      "epoch": 107.57,
+      "grad_norm": 19.16034507751465,
+      "learning_rate": 6.414414414414415e-06,
+      "loss": 0.4482,
+      "step": 35820
+    },
+    {
+      "epoch": 107.6,
+      "grad_norm": 13.354146003723145,
+      "learning_rate": 6.4134134134134145e-06,
+      "loss": 0.4796,
+      "step": 35830
+    },
+    {
+      "epoch": 107.63,
+      "grad_norm": 17.73883056640625,
+      "learning_rate": 6.412412412412413e-06,
+      "loss": 0.4189,
+      "step": 35840
+    },
+    {
+      "epoch": 107.66,
+      "grad_norm": 17.11826515197754,
+      "learning_rate": 6.4114114114114125e-06,
+      "loss": 0.4451,
+      "step": 35850
+    },
+    {
+      "epoch": 107.69,
+      "grad_norm": 9.222671508789062,
+      "learning_rate": 6.410410410410411e-06,
+      "loss": 0.4672,
+      "step": 35860
+    },
+    {
+      "epoch": 107.72,
+      "grad_norm": 14.512218475341797,
+      "learning_rate": 6.40940940940941e-06,
+      "loss": 0.4488,
+      "step": 35870
+    },
+    {
+      "epoch": 107.75,
+      "grad_norm": 14.631092071533203,
+      "learning_rate": 6.408408408408408e-06,
+      "loss": 0.432,
+      "step": 35880
+    },
+    {
+      "epoch": 107.78,
+      "grad_norm": 12.233695983886719,
+      "learning_rate": 6.407407407407408e-06,
+      "loss": 0.4694,
+      "step": 35890
+    },
+    {
+      "epoch": 107.81,
+      "grad_norm": 18.23126220703125,
+      "learning_rate": 6.406406406406407e-06,
+      "loss": 0.503,
+      "step": 35900
+    },
+    {
+      "epoch": 107.84,
+      "grad_norm": 15.379576683044434,
+      "learning_rate": 6.405405405405406e-06,
+      "loss": 0.4538,
+      "step": 35910
+    },
+    {
+      "epoch": 107.87,
+      "grad_norm": 13.152854919433594,
+      "learning_rate": 6.404404404404405e-06,
+      "loss": 0.4198,
+      "step": 35920
+    },
+    {
+      "epoch": 107.9,
+      "grad_norm": 17.65294075012207,
+      "learning_rate": 6.403403403403405e-06,
+      "loss": 0.4661,
+      "step": 35930
+    },
+    {
+      "epoch": 107.93,
+      "grad_norm": 19.87003517150879,
+      "learning_rate": 6.402402402402403e-06,
+      "loss": 0.4846,
+      "step": 35940
+    },
+    {
+      "epoch": 107.96,
+      "grad_norm": 11.539828300476074,
+      "learning_rate": 6.401401401401402e-06,
+      "loss": 0.471,
+      "step": 35950
+    },
+    {
+      "epoch": 107.99,
+      "grad_norm": 18.90138053894043,
+      "learning_rate": 6.4004004004004004e-06,
+      "loss": 0.4358,
+      "step": 35960
+    },
+    {
+      "epoch": 108.0,
+      "eval_accuracy": 0.8643,
+      "eval_loss": 0.4953688383102417,
+      "eval_runtime": 12.7242,
+      "eval_samples_per_second": 785.904,
+      "eval_steps_per_second": 3.144,
+      "step": 35964
+    },
+    {
+      "epoch": 108.02,
+      "grad_norm": 13.45108413696289,
+      "learning_rate": 6.3993993993994e-06,
+      "loss": 0.614,
+      "step": 35970
+    },
+    {
+      "epoch": 108.05,
+      "grad_norm": 14.492265701293945,
+      "learning_rate": 6.3983983983983985e-06,
+      "loss": 0.4587,
+      "step": 35980
+    },
+    {
+      "epoch": 108.08,
+      "grad_norm": 16.008377075195312,
+      "learning_rate": 6.397397397397398e-06,
+      "loss": 0.4769,
+      "step": 35990
+    },
+    {
+      "epoch": 108.11,
+      "grad_norm": 18.540145874023438,
+      "learning_rate": 6.396396396396397e-06,
+      "loss": 0.4383,
+      "step": 36000
+    },
+    {
+      "epoch": 108.14,
+      "grad_norm": 20.01531410217285,
+      "learning_rate": 6.395395395395396e-06,
+      "loss": 0.4609,
+      "step": 36010
+    },
+    {
+      "epoch": 108.17,
+      "grad_norm": 14.863042831420898,
+      "learning_rate": 6.394394394394395e-06,
+      "loss": 0.4309,
+      "step": 36020
+    },
+    {
+      "epoch": 108.2,
+      "grad_norm": 14.092924118041992,
+      "learning_rate": 6.393393393393393e-06,
+      "loss": 0.4305,
+      "step": 36030
+    },
+    {
+      "epoch": 108.23,
+      "grad_norm": 14.197072982788086,
+      "learning_rate": 6.392392392392393e-06,
+      "loss": 0.4745,
+      "step": 36040
+    },
+    {
+      "epoch": 108.26,
+      "grad_norm": 15.748990058898926,
+      "learning_rate": 6.391391391391392e-06,
+      "loss": 0.4954,
+      "step": 36050
+    },
+    {
+      "epoch": 108.29,
+      "grad_norm": 12.143659591674805,
+      "learning_rate": 6.390390390390391e-06,
+      "loss": 0.4426,
+      "step": 36060
+    },
+    {
+      "epoch": 108.32,
+      "grad_norm": 13.76645278930664,
+      "learning_rate": 6.38938938938939e-06,
+      "loss": 0.4132,
+      "step": 36070
+    },
+    {
+      "epoch": 108.35,
+      "grad_norm": 16.62897300720215,
+      "learning_rate": 6.388388388388389e-06,
+      "loss": 0.4624,
+      "step": 36080
+    },
+    {
+      "epoch": 108.38,
+      "grad_norm": 13.602453231811523,
+      "learning_rate": 6.387387387387388e-06,
+      "loss": 0.4503,
+      "step": 36090
+    },
+    {
+      "epoch": 108.41,
+      "grad_norm": 16.384037017822266,
+      "learning_rate": 6.3863863863863875e-06,
+      "loss": 0.4368,
+      "step": 36100
+    },
+    {
+      "epoch": 108.44,
+      "grad_norm": 19.941577911376953,
+      "learning_rate": 6.385385385385386e-06,
+      "loss": 0.4598,
+      "step": 36110
+    },
+    {
+      "epoch": 108.47,
+      "grad_norm": 17.934188842773438,
+      "learning_rate": 6.384384384384385e-06,
+      "loss": 0.4008,
+      "step": 36120
+    },
+    {
+      "epoch": 108.5,
+      "grad_norm": 11.638772010803223,
+      "learning_rate": 6.383383383383383e-06,
+      "loss": 0.493,
+      "step": 36130
+    },
+    {
+      "epoch": 108.53,
+      "grad_norm": 14.956866264343262,
+      "learning_rate": 6.382382382382383e-06,
+      "loss": 0.5107,
+      "step": 36140
+    },
+    {
+      "epoch": 108.56,
+      "grad_norm": 12.474369049072266,
+      "learning_rate": 6.381381381381382e-06,
+      "loss": 0.4774,
+      "step": 36150
+    },
+    {
+      "epoch": 108.59,
+      "grad_norm": 13.798026084899902,
+      "learning_rate": 6.380380380380381e-06,
+      "loss": 0.4786,
+      "step": 36160
+    },
+    {
+      "epoch": 108.62,
+      "grad_norm": 16.548065185546875,
+      "learning_rate": 6.37937937937938e-06,
+      "loss": 0.4017,
+      "step": 36170
+    },
+    {
+      "epoch": 108.65,
+      "grad_norm": 12.218683242797852,
+      "learning_rate": 6.378378378378379e-06,
+      "loss": 0.4989,
+      "step": 36180
+    },
+    {
+      "epoch": 108.68,
+      "grad_norm": 18.878925323486328,
+      "learning_rate": 6.377377377377378e-06,
+      "loss": 0.4649,
+      "step": 36190
+    },
+    {
+      "epoch": 108.71,
+      "grad_norm": 11.830445289611816,
+      "learning_rate": 6.376376376376376e-06,
+      "loss": 0.4534,
+      "step": 36200
+    },
+    {
+      "epoch": 108.74,
+      "grad_norm": 13.8546781539917,
+      "learning_rate": 6.3753753753753755e-06,
+      "loss": 0.4796,
+      "step": 36210
+    },
+    {
+      "epoch": 108.77,
+      "grad_norm": 15.33486270904541,
+      "learning_rate": 6.374374374374375e-06,
+      "loss": 0.4753,
+      "step": 36220
+    },
+    {
+      "epoch": 108.8,
+      "grad_norm": 11.009231567382812,
+      "learning_rate": 6.3733733733733735e-06,
+      "loss": 0.4376,
+      "step": 36230
+    },
+    {
+      "epoch": 108.83,
+      "grad_norm": 12.35025405883789,
+      "learning_rate": 6.372372372372373e-06,
+      "loss": 0.4757,
+      "step": 36240
+    },
+    {
+      "epoch": 108.86,
+      "grad_norm": 15.227595329284668,
+      "learning_rate": 6.371371371371372e-06,
+      "loss": 0.4199,
+      "step": 36250
+    },
+    {
+      "epoch": 108.89,
+      "grad_norm": 14.35750675201416,
+      "learning_rate": 6.370370370370371e-06,
+      "loss": 0.4686,
+      "step": 36260
+    },
+    {
+      "epoch": 108.92,
+      "grad_norm": 12.59177303314209,
+      "learning_rate": 6.3693693693693704e-06,
+      "loss": 0.4496,
+      "step": 36270
+    },
+    {
+      "epoch": 108.95,
+      "grad_norm": 15.565778732299805,
+      "learning_rate": 6.368368368368368e-06,
+      "loss": 0.5008,
+      "step": 36280
+    },
+    {
+      "epoch": 108.98,
+      "grad_norm": 10.969858169555664,
+      "learning_rate": 6.367367367367368e-06,
+      "loss": 0.4588,
+      "step": 36290
+    },
+    {
+      "epoch": 109.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.4922635853290558,
+      "eval_runtime": 12.9316,
+      "eval_samples_per_second": 773.298,
+      "eval_steps_per_second": 3.093,
+      "step": 36297
+    },
+    {
+      "epoch": 109.01,
+      "grad_norm": 14.826801300048828,
+      "learning_rate": 6.366366366366366e-06,
+      "loss": 0.5274,
+      "step": 36300
+    },
+    {
+      "epoch": 109.04,
+      "grad_norm": 11.0240478515625,
+      "learning_rate": 6.365365365365366e-06,
+      "loss": 0.4074,
+      "step": 36310
+    },
+    {
+      "epoch": 109.07,
+      "grad_norm": 17.72291374206543,
+      "learning_rate": 6.364364364364365e-06,
+      "loss": 0.4299,
+      "step": 36320
+    },
+    {
+      "epoch": 109.1,
+      "grad_norm": 16.200233459472656,
+      "learning_rate": 6.363363363363364e-06,
+      "loss": 0.4693,
+      "step": 36330
+    },
+    {
+      "epoch": 109.13,
+      "grad_norm": 17.902406692504883,
+      "learning_rate": 6.362362362362363e-06,
+      "loss": 0.4359,
+      "step": 36340
+    },
+    {
+      "epoch": 109.16,
+      "grad_norm": 11.185669898986816,
+      "learning_rate": 6.361361361361363e-06,
+      "loss": 0.4448,
+      "step": 36350
+    },
+    {
+      "epoch": 109.19,
+      "grad_norm": 11.728100776672363,
+      "learning_rate": 6.360360360360361e-06,
+      "loss": 0.4523,
+      "step": 36360
+    },
+    {
+      "epoch": 109.22,
+      "grad_norm": 15.993358612060547,
+      "learning_rate": 6.35935935935936e-06,
+      "loss": 0.4684,
+      "step": 36370
+    },
+    {
+      "epoch": 109.25,
+      "grad_norm": 20.444183349609375,
+      "learning_rate": 6.358358358358358e-06,
+      "loss": 0.429,
+      "step": 36380
+    },
+    {
+      "epoch": 109.28,
+      "grad_norm": 10.875001907348633,
+      "learning_rate": 6.357357357357358e-06,
+      "loss": 0.4715,
+      "step": 36390
+    },
+    {
+      "epoch": 109.31,
+      "grad_norm": 13.275202751159668,
+      "learning_rate": 6.356356356356357e-06,
+      "loss": 0.4605,
+      "step": 36400
+    },
+    {
+      "epoch": 109.34,
+      "grad_norm": 14.225875854492188,
+      "learning_rate": 6.355355355355356e-06,
+      "loss": 0.4239,
+      "step": 36410
+    },
+    {
+      "epoch": 109.37,
+      "grad_norm": 18.543405532836914,
+      "learning_rate": 6.354354354354355e-06,
+      "loss": 0.4541,
+      "step": 36420
+    },
+    {
+      "epoch": 109.4,
+      "grad_norm": 12.068740844726562,
+      "learning_rate": 6.353353353353354e-06,
+      "loss": 0.4537,
+      "step": 36430
+    },
+    {
+      "epoch": 109.43,
+      "grad_norm": 16.46315574645996,
+      "learning_rate": 6.352352352352353e-06,
+      "loss": 0.442,
+      "step": 36440
+    },
+    {
+      "epoch": 109.46,
+      "grad_norm": 16.454023361206055,
+      "learning_rate": 6.351351351351351e-06,
+      "loss": 0.4253,
+      "step": 36450
+    },
+    {
+      "epoch": 109.49,
+      "grad_norm": 13.146560668945312,
+      "learning_rate": 6.3503503503503505e-06,
+      "loss": 0.4535,
+      "step": 36460
+    },
+    {
+      "epoch": 109.52,
+      "grad_norm": 24.77417755126953,
+      "learning_rate": 6.34934934934935e-06,
+      "loss": 0.4464,
+      "step": 36470
+    },
+    {
+      "epoch": 109.55,
+      "grad_norm": 18.188602447509766,
+      "learning_rate": 6.3483483483483485e-06,
+      "loss": 0.5332,
+      "step": 36480
+    },
+    {
+      "epoch": 109.58,
+      "grad_norm": 12.230823516845703,
+      "learning_rate": 6.347347347347348e-06,
+      "loss": 0.4418,
+      "step": 36490
+    },
+    {
+      "epoch": 109.61,
+      "grad_norm": 13.86673641204834,
+      "learning_rate": 6.3463463463463474e-06,
+      "loss": 0.4938,
+      "step": 36500
+    },
+    {
+      "epoch": 109.64,
+      "grad_norm": 14.99777889251709,
+      "learning_rate": 6.345345345345346e-06,
+      "loss": 0.4442,
+      "step": 36510
+    },
+    {
+      "epoch": 109.67,
+      "grad_norm": 15.082866668701172,
+      "learning_rate": 6.3443443443443455e-06,
+      "loss": 0.4255,
+      "step": 36520
+    },
+    {
+      "epoch": 109.7,
+      "grad_norm": 13.000606536865234,
+      "learning_rate": 6.343343343343343e-06,
+      "loss": 0.4327,
+      "step": 36530
+    },
+    {
+      "epoch": 109.73,
+      "grad_norm": 16.731948852539062,
+      "learning_rate": 6.342342342342343e-06,
+      "loss": 0.4619,
+      "step": 36540
+    },
+    {
+      "epoch": 109.76,
+      "grad_norm": 13.189562797546387,
+      "learning_rate": 6.341341341341341e-06,
+      "loss": 0.3816,
+      "step": 36550
+    },
+    {
+      "epoch": 109.79,
+      "grad_norm": 28.662769317626953,
+      "learning_rate": 6.340340340340341e-06,
+      "loss": 0.4278,
+      "step": 36560
+    },
+    {
+      "epoch": 109.82,
+      "grad_norm": 13.81191349029541,
+      "learning_rate": 6.33933933933934e-06,
+      "loss": 0.4628,
+      "step": 36570
+    },
+    {
+      "epoch": 109.85,
+      "grad_norm": 27.256855010986328,
+      "learning_rate": 6.338338338338339e-06,
+      "loss": 0.4455,
+      "step": 36580
+    },
+    {
+      "epoch": 109.88,
+      "grad_norm": 12.601924896240234,
+      "learning_rate": 6.337337337337338e-06,
+      "loss": 0.4463,
+      "step": 36590
+    },
+    {
+      "epoch": 109.91,
+      "grad_norm": 28.672277450561523,
+      "learning_rate": 6.336336336336338e-06,
+      "loss": 0.5593,
+      "step": 36600
+    },
+    {
+      "epoch": 109.94,
+      "grad_norm": 19.5489559173584,
+      "learning_rate": 6.335335335335336e-06,
+      "loss": 0.4056,
+      "step": 36610
+    },
+    {
+      "epoch": 109.97,
+      "grad_norm": 10.736978530883789,
+      "learning_rate": 6.334334334334335e-06,
+      "loss": 0.422,
+      "step": 36620
+    },
+    {
+      "epoch": 110.0,
+      "grad_norm": 102.30081939697266,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.4681,
+      "step": 36630
+    },
+    {
+      "epoch": 110.0,
+      "eval_accuracy": 0.8654,
+      "eval_loss": 0.4947918951511383,
+      "eval_runtime": 12.4535,
+      "eval_samples_per_second": 802.988,
+      "eval_steps_per_second": 3.212,
+      "step": 36630
+    },
+    {
+      "epoch": 110.03,
+      "grad_norm": 14.615818977355957,
+      "learning_rate": 6.332332332332333e-06,
+      "loss": 0.4471,
+      "step": 36640
+    },
+    {
+      "epoch": 110.06,
+      "grad_norm": 15.137236595153809,
+      "learning_rate": 6.3313313313313314e-06,
+      "loss": 0.4622,
+      "step": 36650
+    },
+    {
+      "epoch": 110.09,
+      "grad_norm": 13.666529655456543,
+      "learning_rate": 6.330330330330331e-06,
+      "loss": 0.4345,
+      "step": 36660
+    },
+    {
+      "epoch": 110.12,
+      "grad_norm": 17.314367294311523,
+      "learning_rate": 6.32932932932933e-06,
+      "loss": 0.5547,
+      "step": 36670
+    },
+    {
+      "epoch": 110.15,
+      "grad_norm": 17.806934356689453,
+      "learning_rate": 6.328328328328329e-06,
+      "loss": 0.4629,
+      "step": 36680
+    },
+    {
+      "epoch": 110.18,
+      "grad_norm": 22.68148422241211,
+      "learning_rate": 6.327327327327328e-06,
+      "loss": 0.4405,
+      "step": 36690
+    },
+    {
+      "epoch": 110.21,
+      "grad_norm": 14.4562406539917,
+      "learning_rate": 6.326326326326326e-06,
+      "loss": 0.4969,
+      "step": 36700
+    },
+    {
+      "epoch": 110.24,
+      "grad_norm": 16.531978607177734,
+      "learning_rate": 6.3253253253253255e-06,
+      "loss": 0.4643,
+      "step": 36710
+    },
+    {
+      "epoch": 110.27,
+      "grad_norm": 12.48605728149414,
+      "learning_rate": 6.324324324324325e-06,
+      "loss": 0.4613,
+      "step": 36720
+    },
+    {
+      "epoch": 110.3,
+      "grad_norm": 18.395578384399414,
+      "learning_rate": 6.323323323323324e-06,
+      "loss": 0.4706,
+      "step": 36730
+    },
+    {
+      "epoch": 110.33,
+      "grad_norm": 13.167364120483398,
+      "learning_rate": 6.322322322322323e-06,
+      "loss": 0.4661,
+      "step": 36740
+    },
+    {
+      "epoch": 110.36,
+      "grad_norm": 15.323986053466797,
+      "learning_rate": 6.321321321321322e-06,
+      "loss": 0.4234,
+      "step": 36750
+    },
+    {
+      "epoch": 110.39,
+      "grad_norm": 13.639139175415039,
+      "learning_rate": 6.320320320320321e-06,
+      "loss": 0.432,
+      "step": 36760
+    },
+    {
+      "epoch": 110.42,
+      "grad_norm": 14.6893892288208,
+      "learning_rate": 6.3193193193193205e-06,
+      "loss": 0.4591,
+      "step": 36770
+    },
+    {
+      "epoch": 110.45,
+      "grad_norm": 10.424907684326172,
+      "learning_rate": 6.318318318318318e-06,
+      "loss": 0.4808,
+      "step": 36780
+    },
+    {
+      "epoch": 110.48,
+      "grad_norm": 14.365670204162598,
+      "learning_rate": 6.317317317317318e-06,
+      "loss": 0.4844,
+      "step": 36790
+    },
+    {
+      "epoch": 110.51,
+      "grad_norm": 12.55785846710205,
+      "learning_rate": 6.316316316316316e-06,
+      "loss": 0.4757,
+      "step": 36800
+    },
+    {
+      "epoch": 110.54,
+      "grad_norm": 16.355621337890625,
+      "learning_rate": 6.315315315315316e-06,
+      "loss": 0.4082,
+      "step": 36810
+    },
+    {
+      "epoch": 110.57,
+      "grad_norm": 15.718585968017578,
+      "learning_rate": 6.314314314314315e-06,
+      "loss": 0.4355,
+      "step": 36820
+    },
+    {
+      "epoch": 110.6,
+      "grad_norm": 16.67154312133789,
+      "learning_rate": 6.313313313313314e-06,
+      "loss": 0.4768,
+      "step": 36830
+    },
+    {
+      "epoch": 110.63,
+      "grad_norm": 15.919441223144531,
+      "learning_rate": 6.312312312312313e-06,
+      "loss": 0.4522,
+      "step": 36840
+    },
+    {
+      "epoch": 110.66,
+      "grad_norm": 17.336137771606445,
+      "learning_rate": 6.311311311311313e-06,
+      "loss": 0.4044,
+      "step": 36850
+    },
+    {
+      "epoch": 110.69,
+      "grad_norm": 12.74454402923584,
+      "learning_rate": 6.31031031031031e-06,
+      "loss": 0.4595,
+      "step": 36860
+    },
+    {
+      "epoch": 110.72,
+      "grad_norm": 14.225423812866211,
+      "learning_rate": 6.309309309309309e-06,
+      "loss": 0.3993,
+      "step": 36870
+    },
+    {
+      "epoch": 110.75,
+      "grad_norm": 15.219978332519531,
+      "learning_rate": 6.3083083083083084e-06,
+      "loss": 0.4614,
+      "step": 36880
+    },
+    {
+      "epoch": 110.78,
+      "grad_norm": 14.711874961853027,
+      "learning_rate": 6.307307307307308e-06,
+      "loss": 0.4291,
+      "step": 36890
+    },
+    {
+      "epoch": 110.81,
+      "grad_norm": 23.210309982299805,
+      "learning_rate": 6.3063063063063065e-06,
+      "loss": 0.5008,
+      "step": 36900
+    },
+    {
+      "epoch": 110.84,
+      "grad_norm": 24.133777618408203,
+      "learning_rate": 6.305305305305306e-06,
+      "loss": 0.4196,
+      "step": 36910
+    },
+    {
+      "epoch": 110.87,
+      "grad_norm": 14.790199279785156,
+      "learning_rate": 6.304304304304305e-06,
+      "loss": 0.4634,
+      "step": 36920
+    },
+    {
+      "epoch": 110.9,
+      "grad_norm": 9.704394340515137,
+      "learning_rate": 6.303303303303304e-06,
+      "loss": 0.5303,
+      "step": 36930
+    },
+    {
+      "epoch": 110.93,
+      "grad_norm": 12.200167655944824,
+      "learning_rate": 6.302302302302303e-06,
+      "loss": 0.4158,
+      "step": 36940
+    },
+    {
+      "epoch": 110.96,
+      "grad_norm": 16.717378616333008,
+      "learning_rate": 6.301301301301301e-06,
+      "loss": 0.4564,
+      "step": 36950
+    },
+    {
+      "epoch": 110.99,
+      "grad_norm": 17.715986251831055,
+      "learning_rate": 6.300300300300301e-06,
+      "loss": 0.4602,
+      "step": 36960
+    },
+    {
+      "epoch": 111.0,
+      "eval_accuracy": 0.8677,
+      "eval_loss": 0.4961108863353729,
+      "eval_runtime": 13.1448,
+      "eval_samples_per_second": 760.755,
+      "eval_steps_per_second": 3.043,
+      "step": 36963
+    },
+    {
+      "epoch": 111.02,
+      "grad_norm": 16.29159927368164,
+      "learning_rate": 6.2992992992993e-06,
+      "loss": 0.4185,
+      "step": 36970
+    },
+    {
+      "epoch": 111.05,
+      "grad_norm": 18.87250328063965,
+      "learning_rate": 6.298298298298299e-06,
+      "loss": 0.4034,
+      "step": 36980
+    },
+    {
+      "epoch": 111.08,
+      "grad_norm": 16.234193801879883,
+      "learning_rate": 6.297297297297298e-06,
+      "loss": 0.4133,
+      "step": 36990
+    },
+    {
+      "epoch": 111.11,
+      "grad_norm": 11.578213691711426,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.4355,
+      "step": 37000
+    },
+    {
+      "epoch": 111.14,
+      "grad_norm": 16.66002655029297,
+      "learning_rate": 6.295295295295296e-06,
+      "loss": 0.4949,
+      "step": 37010
+    },
+    {
+      "epoch": 111.17,
+      "grad_norm": 16.080270767211914,
+      "learning_rate": 6.2942942942942955e-06,
+      "loss": 0.448,
+      "step": 37020
+    },
+    {
+      "epoch": 111.2,
+      "grad_norm": 14.625487327575684,
+      "learning_rate": 6.293293293293293e-06,
+      "loss": 0.3734,
+      "step": 37030
+    },
+    {
+      "epoch": 111.23,
+      "grad_norm": 14.804281234741211,
+      "learning_rate": 6.292292292292293e-06,
+      "loss": 0.3555,
+      "step": 37040
+    },
+    {
+      "epoch": 111.26,
+      "grad_norm": 21.609233856201172,
+      "learning_rate": 6.291291291291291e-06,
+      "loss": 0.4371,
+      "step": 37050
+    },
+    {
+      "epoch": 111.29,
+      "grad_norm": 17.855134963989258,
+      "learning_rate": 6.290290290290291e-06,
+      "loss": 0.4519,
+      "step": 37060
+    },
+    {
+      "epoch": 111.32,
+      "grad_norm": 14.120532989501953,
+      "learning_rate": 6.28928928928929e-06,
+      "loss": 0.3552,
+      "step": 37070
+    },
+    {
+      "epoch": 111.35,
+      "grad_norm": 16.431406021118164,
+      "learning_rate": 6.288288288288289e-06,
+      "loss": 0.4256,
+      "step": 37080
+    },
+    {
+      "epoch": 111.38,
+      "grad_norm": 21.464092254638672,
+      "learning_rate": 6.287287287287288e-06,
+      "loss": 0.4372,
+      "step": 37090
+    },
+    {
+      "epoch": 111.41,
+      "grad_norm": 13.230178833007812,
+      "learning_rate": 6.286286286286287e-06,
+      "loss": 0.4625,
+      "step": 37100
+    },
+    {
+      "epoch": 111.44,
+      "grad_norm": 12.773159980773926,
+      "learning_rate": 6.2852852852852854e-06,
+      "loss": 0.4094,
+      "step": 37110
+    },
+    {
+      "epoch": 111.47,
+      "grad_norm": 15.667949676513672,
+      "learning_rate": 6.284284284284284e-06,
+      "loss": 0.4193,
+      "step": 37120
+    },
+    {
+      "epoch": 111.5,
+      "grad_norm": 16.480850219726562,
+      "learning_rate": 6.2832832832832835e-06,
+      "loss": 0.4253,
+      "step": 37130
+    },
+    {
+      "epoch": 111.53,
+      "grad_norm": 19.231075286865234,
+      "learning_rate": 6.282282282282283e-06,
+      "loss": 0.3951,
+      "step": 37140
+    },
+    {
+      "epoch": 111.56,
+      "grad_norm": 20.724079132080078,
+      "learning_rate": 6.2812812812812815e-06,
+      "loss": 0.4371,
+      "step": 37150
+    },
+    {
+      "epoch": 111.59,
+      "grad_norm": 16.201953887939453,
+      "learning_rate": 6.280280280280281e-06,
+      "loss": 0.465,
+      "step": 37160
+    },
+    {
+      "epoch": 111.62,
+      "grad_norm": 15.528708457946777,
+      "learning_rate": 6.27927927927928e-06,
+      "loss": 0.471,
+      "step": 37170
+    },
+    {
+      "epoch": 111.65,
+      "grad_norm": 15.826865196228027,
+      "learning_rate": 6.278278278278279e-06,
+      "loss": 0.4455,
+      "step": 37180
+    },
+    {
+      "epoch": 111.68,
+      "grad_norm": 19.372074127197266,
+      "learning_rate": 6.2772772772772784e-06,
+      "loss": 0.5016,
+      "step": 37190
+    },
+    {
+      "epoch": 111.71,
+      "grad_norm": 29.868099212646484,
+      "learning_rate": 6.276276276276276e-06,
+      "loss": 0.4697,
+      "step": 37200
+    },
+    {
+      "epoch": 111.74,
+      "grad_norm": 23.138473510742188,
+      "learning_rate": 6.275275275275276e-06,
+      "loss": 0.4168,
+      "step": 37210
+    },
+    {
+      "epoch": 111.77,
+      "grad_norm": 14.392965316772461,
+      "learning_rate": 6.274274274274274e-06,
+      "loss": 0.4634,
+      "step": 37220
+    },
+    {
+      "epoch": 111.8,
+      "grad_norm": 12.71838092803955,
+      "learning_rate": 6.273273273273274e-06,
+      "loss": 0.5002,
+      "step": 37230
+    },
+    {
+      "epoch": 111.83,
+      "grad_norm": 12.515597343444824,
+      "learning_rate": 6.272272272272273e-06,
+      "loss": 0.4426,
+      "step": 37240
+    },
+    {
+      "epoch": 111.86,
+      "grad_norm": 13.296930313110352,
+      "learning_rate": 6.271271271271272e-06,
+      "loss": 0.4846,
+      "step": 37250
+    },
+    {
+      "epoch": 111.89,
+      "grad_norm": 9.491498947143555,
+      "learning_rate": 6.270270270270271e-06,
+      "loss": 0.4136,
+      "step": 37260
+    },
+    {
+      "epoch": 111.92,
+      "grad_norm": 22.016929626464844,
+      "learning_rate": 6.2692692692692706e-06,
+      "loss": 0.4664,
+      "step": 37270
+    },
+    {
+      "epoch": 111.95,
+      "grad_norm": 10.204870223999023,
+      "learning_rate": 6.268268268268268e-06,
+      "loss": 0.4187,
+      "step": 37280
+    },
+    {
+      "epoch": 111.98,
+      "grad_norm": 17.137714385986328,
+      "learning_rate": 6.267267267267268e-06,
+      "loss": 0.4871,
+      "step": 37290
+    },
+    {
+      "epoch": 112.0,
+      "eval_accuracy": 0.8634,
+      "eval_loss": 0.5004977583885193,
+      "eval_runtime": 12.9395,
+      "eval_samples_per_second": 772.825,
+      "eval_steps_per_second": 3.091,
+      "step": 37296
+    },
+    {
+      "epoch": 112.01,
+      "grad_norm": 15.295434951782227,
+      "learning_rate": 6.266266266266266e-06,
+      "loss": 0.4445,
+      "step": 37300
+    },
+    {
+      "epoch": 112.04,
+      "grad_norm": 15.803238868713379,
+      "learning_rate": 6.265265265265266e-06,
+      "loss": 0.4706,
+      "step": 37310
+    },
+    {
+      "epoch": 112.07,
+      "grad_norm": 24.0127010345459,
+      "learning_rate": 6.264264264264264e-06,
+      "loss": 0.3917,
+      "step": 37320
+    },
+    {
+      "epoch": 112.1,
+      "grad_norm": 18.407052993774414,
+      "learning_rate": 6.263263263263264e-06,
+      "loss": 0.4794,
+      "step": 37330
+    },
+    {
+      "epoch": 112.13,
+      "grad_norm": 8.669754981994629,
+      "learning_rate": 6.262262262262263e-06,
+      "loss": 0.4547,
+      "step": 37340
+    },
+    {
+      "epoch": 112.16,
+      "grad_norm": 12.482906341552734,
+      "learning_rate": 6.261261261261262e-06,
+      "loss": 0.4149,
+      "step": 37350
+    },
+    {
+      "epoch": 112.19,
+      "grad_norm": 22.047351837158203,
+      "learning_rate": 6.2602602602602605e-06,
+      "loss": 0.4477,
+      "step": 37360
+    },
+    {
+      "epoch": 112.22,
+      "grad_norm": 17.21535873413086,
+      "learning_rate": 6.259259259259259e-06,
+      "loss": 0.423,
+      "step": 37370
+    },
+    {
+      "epoch": 112.25,
+      "grad_norm": 16.480724334716797,
+      "learning_rate": 6.2582582582582585e-06,
+      "loss": 0.4289,
+      "step": 37380
+    },
+    {
+      "epoch": 112.28,
+      "grad_norm": 18.517776489257812,
+      "learning_rate": 6.257257257257258e-06,
+      "loss": 0.5141,
+      "step": 37390
+    },
+    {
+      "epoch": 112.31,
+      "grad_norm": 18.755706787109375,
+      "learning_rate": 6.2562562562562565e-06,
+      "loss": 0.4712,
+      "step": 37400
+    },
+    {
+      "epoch": 112.34,
+      "grad_norm": 12.167431831359863,
+      "learning_rate": 6.255255255255256e-06,
+      "loss": 0.4325,
+      "step": 37410
+    },
+    {
+      "epoch": 112.37,
+      "grad_norm": 23.029727935791016,
+      "learning_rate": 6.2542542542542554e-06,
+      "loss": 0.4161,
+      "step": 37420
+    },
+    {
+      "epoch": 112.4,
+      "grad_norm": 13.140212059020996,
+      "learning_rate": 6.253253253253254e-06,
+      "loss": 0.4778,
+      "step": 37430
+    },
+    {
+      "epoch": 112.43,
+      "grad_norm": 12.20968246459961,
+      "learning_rate": 6.2522522522522535e-06,
+      "loss": 0.4465,
+      "step": 37440
+    },
+    {
+      "epoch": 112.46,
+      "grad_norm": 11.349285125732422,
+      "learning_rate": 6.251251251251251e-06,
+      "loss": 0.4764,
+      "step": 37450
+    },
+    {
+      "epoch": 112.49,
+      "grad_norm": 19.511850357055664,
+      "learning_rate": 6.250250250250251e-06,
+      "loss": 0.4146,
+      "step": 37460
+    },
+    {
+      "epoch": 112.52,
+      "grad_norm": 17.5305233001709,
+      "learning_rate": 6.249249249249249e-06,
+      "loss": 0.4457,
+      "step": 37470
+    },
+    {
+      "epoch": 112.55,
+      "grad_norm": 12.776636123657227,
+      "learning_rate": 6.248248248248249e-06,
+      "loss": 0.4541,
+      "step": 37480
+    },
+    {
+      "epoch": 112.58,
+      "grad_norm": 15.932623863220215,
+      "learning_rate": 6.247247247247248e-06,
+      "loss": 0.3982,
+      "step": 37490
+    },
+    {
+      "epoch": 112.61,
+      "grad_norm": 15.5503511428833,
+      "learning_rate": 6.246246246246247e-06,
+      "loss": 0.4672,
+      "step": 37500
+    },
+    {
+      "epoch": 112.64,
+      "grad_norm": 17.59156608581543,
+      "learning_rate": 6.245245245245246e-06,
+      "loss": 0.4098,
+      "step": 37510
+    },
+    {
+      "epoch": 112.67,
+      "grad_norm": 16.5859317779541,
+      "learning_rate": 6.244244244244246e-06,
+      "loss": 0.4281,
+      "step": 37520
+    },
+    {
+      "epoch": 112.7,
+      "grad_norm": 18.925283432006836,
+      "learning_rate": 6.243243243243243e-06,
+      "loss": 0.4565,
+      "step": 37530
+    },
+    {
+      "epoch": 112.73,
+      "grad_norm": 19.63197898864746,
+      "learning_rate": 6.242242242242243e-06,
+      "loss": 0.4092,
+      "step": 37540
+    },
+    {
+      "epoch": 112.76,
+      "grad_norm": 9.282973289489746,
+      "learning_rate": 6.241241241241241e-06,
+      "loss": 0.4131,
+      "step": 37550
+    },
+    {
+      "epoch": 112.79,
+      "grad_norm": 23.554494857788086,
+      "learning_rate": 6.240240240240241e-06,
+      "loss": 0.4367,
+      "step": 37560
+    },
+    {
+      "epoch": 112.82,
+      "grad_norm": 9.415934562683105,
+      "learning_rate": 6.2392392392392394e-06,
+      "loss": 0.4277,
+      "step": 37570
+    },
+    {
+      "epoch": 112.85,
+      "grad_norm": 16.967266082763672,
+      "learning_rate": 6.238238238238239e-06,
+      "loss": 0.4318,
+      "step": 37580
+    },
+    {
+      "epoch": 112.88,
+      "grad_norm": 14.02496337890625,
+      "learning_rate": 6.237237237237238e-06,
+      "loss": 0.4207,
+      "step": 37590
+    },
+    {
+      "epoch": 112.91,
+      "grad_norm": 16.25889015197754,
+      "learning_rate": 6.236236236236237e-06,
+      "loss": 0.4534,
+      "step": 37600
+    },
+    {
+      "epoch": 112.94,
+      "grad_norm": 15.557313919067383,
+      "learning_rate": 6.2352352352352355e-06,
+      "loss": 0.422,
+      "step": 37610
+    },
+    {
+      "epoch": 112.97,
+      "grad_norm": 12.834463119506836,
+      "learning_rate": 6.234234234234234e-06,
+      "loss": 0.4144,
+      "step": 37620
+    },
+    {
+      "epoch": 113.0,
+      "eval_accuracy": 0.8657,
+      "eval_loss": 0.49875473976135254,
+      "eval_runtime": 12.5223,
+      "eval_samples_per_second": 798.577,
+      "eval_steps_per_second": 3.194,
+      "step": 37629
+    },
+    {
+      "epoch": 113.0,
+      "grad_norm": 19.846052169799805,
+      "learning_rate": 6.2332332332332335e-06,
+      "loss": 0.4693,
+      "step": 37630
+    },
+    {
+      "epoch": 113.03,
+      "grad_norm": 19.653499603271484,
+      "learning_rate": 6.232232232232233e-06,
+      "loss": 0.4498,
+      "step": 37640
+    },
+    {
+      "epoch": 113.06,
+      "grad_norm": 36.175758361816406,
+      "learning_rate": 6.2312312312312316e-06,
+      "loss": 0.4219,
+      "step": 37650
+    },
+    {
+      "epoch": 113.09,
+      "grad_norm": 12.447442054748535,
+      "learning_rate": 6.230230230230231e-06,
+      "loss": 0.4807,
+      "step": 37660
+    },
+    {
+      "epoch": 113.12,
+      "grad_norm": 22.12033462524414,
+      "learning_rate": 6.22922922922923e-06,
+      "loss": 0.468,
+      "step": 37670
+    },
+    {
+      "epoch": 113.15,
+      "grad_norm": 13.504697799682617,
+      "learning_rate": 6.228228228228229e-06,
+      "loss": 0.4186,
+      "step": 37680
+    },
+    {
+      "epoch": 113.18,
+      "grad_norm": 19.08098793029785,
+      "learning_rate": 6.2272272272272285e-06,
+      "loss": 0.4318,
+      "step": 37690
+    },
+    {
+      "epoch": 113.21,
+      "grad_norm": 24.229116439819336,
+      "learning_rate": 6.226226226226226e-06,
+      "loss": 0.4485,
+      "step": 37700
+    },
+    {
+      "epoch": 113.24,
+      "grad_norm": 25.67328643798828,
+      "learning_rate": 6.225225225225226e-06,
+      "loss": 0.4525,
+      "step": 37710
+    },
+    {
+      "epoch": 113.27,
+      "grad_norm": 13.256377220153809,
+      "learning_rate": 6.224224224224224e-06,
+      "loss": 0.4272,
+      "step": 37720
+    },
+    {
+      "epoch": 113.3,
+      "grad_norm": 25.311668395996094,
+      "learning_rate": 6.223223223223224e-06,
+      "loss": 0.4347,
+      "step": 37730
+    },
+    {
+      "epoch": 113.33,
+      "grad_norm": 30.365467071533203,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 0.4622,
+      "step": 37740
+    },
+    {
+      "epoch": 113.36,
+      "grad_norm": 14.892073631286621,
+      "learning_rate": 6.221221221221222e-06,
+      "loss": 0.4362,
+      "step": 37750
+    },
+    {
+      "epoch": 113.39,
+      "grad_norm": 12.964187622070312,
+      "learning_rate": 6.220220220220221e-06,
+      "loss": 0.4599,
+      "step": 37760
+    },
+    {
+      "epoch": 113.42,
+      "grad_norm": 13.354842185974121,
+      "learning_rate": 6.219219219219221e-06,
+      "loss": 0.3993,
+      "step": 37770
+    },
+    {
+      "epoch": 113.45,
+      "grad_norm": 16.26036262512207,
+      "learning_rate": 6.218218218218218e-06,
+      "loss": 0.4052,
+      "step": 37780
+    },
+    {
+      "epoch": 113.48,
+      "grad_norm": 15.502280235290527,
+      "learning_rate": 6.217217217217217e-06,
+      "loss": 0.4409,
+      "step": 37790
+    },
+    {
+      "epoch": 113.51,
+      "grad_norm": 18.03598976135254,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.415,
+      "step": 37800
+    },
+    {
+      "epoch": 113.54,
+      "grad_norm": 11.024408340454102,
+      "learning_rate": 6.215215215215216e-06,
+      "loss": 0.3826,
+      "step": 37810
+    },
+    {
+      "epoch": 113.57,
+      "grad_norm": 11.059122085571289,
+      "learning_rate": 6.2142142142142145e-06,
+      "loss": 0.4138,
+      "step": 37820
+    },
+    {
+      "epoch": 113.6,
+      "grad_norm": 14.920058250427246,
+      "learning_rate": 6.213213213213214e-06,
+      "loss": 0.4385,
+      "step": 37830
+    },
+    {
+      "epoch": 113.63,
+      "grad_norm": 17.982412338256836,
+      "learning_rate": 6.212212212212213e-06,
+      "loss": 0.3999,
+      "step": 37840
+    },
+    {
+      "epoch": 113.66,
+      "grad_norm": 10.935893058776855,
+      "learning_rate": 6.211211211211212e-06,
+      "loss": 0.4039,
+      "step": 37850
+    },
+    {
+      "epoch": 113.69,
+      "grad_norm": 13.610403060913086,
+      "learning_rate": 6.2102102102102105e-06,
+      "loss": 0.4495,
+      "step": 37860
+    },
+    {
+      "epoch": 113.72,
+      "grad_norm": 25.733505249023438,
+      "learning_rate": 6.209209209209209e-06,
+      "loss": 0.3807,
+      "step": 37870
+    },
+    {
+      "epoch": 113.75,
+      "grad_norm": 11.191322326660156,
+      "learning_rate": 6.2082082082082086e-06,
+      "loss": 0.4388,
+      "step": 37880
+    },
+    {
+      "epoch": 113.78,
+      "grad_norm": 17.699485778808594,
+      "learning_rate": 6.207207207207208e-06,
+      "loss": 0.4296,
+      "step": 37890
+    },
+    {
+      "epoch": 113.81,
+      "grad_norm": 18.78080177307129,
+      "learning_rate": 6.206206206206207e-06,
+      "loss": 0.4543,
+      "step": 37900
+    },
+    {
+      "epoch": 113.84,
+      "grad_norm": 15.276420593261719,
+      "learning_rate": 6.205205205205206e-06,
+      "loss": 0.4019,
+      "step": 37910
+    },
+    {
+      "epoch": 113.87,
+      "grad_norm": 11.932049751281738,
+      "learning_rate": 6.204204204204205e-06,
+      "loss": 0.4618,
+      "step": 37920
+    },
+    {
+      "epoch": 113.9,
+      "grad_norm": 9.300352096557617,
+      "learning_rate": 6.203203203203204e-06,
+      "loss": 0.4805,
+      "step": 37930
+    },
+    {
+      "epoch": 113.93,
+      "grad_norm": 17.455055236816406,
+      "learning_rate": 6.2022022022022035e-06,
+      "loss": 0.4179,
+      "step": 37940
+    },
+    {
+      "epoch": 113.96,
+      "grad_norm": 20.1407470703125,
+      "learning_rate": 6.201201201201201e-06,
+      "loss": 0.4492,
+      "step": 37950
+    },
+    {
+      "epoch": 113.99,
+      "grad_norm": 10.844948768615723,
+      "learning_rate": 6.200200200200201e-06,
+      "loss": 0.4735,
+      "step": 37960
+    },
+    {
+      "epoch": 114.0,
+      "eval_accuracy": 0.8654,
+      "eval_loss": 0.49756744503974915,
+      "eval_runtime": 13.0591,
+      "eval_samples_per_second": 765.752,
+      "eval_steps_per_second": 3.063,
+      "step": 37962
+    },
+    {
+      "epoch": 114.02,
+      "grad_norm": 12.441977500915527,
+      "learning_rate": 6.199199199199199e-06,
+      "loss": 0.3877,
+      "step": 37970
+    },
+    {
+      "epoch": 114.05,
+      "grad_norm": 15.719921112060547,
+      "learning_rate": 6.198198198198199e-06,
+      "loss": 0.4527,
+      "step": 37980
+    },
+    {
+      "epoch": 114.08,
+      "grad_norm": 16.609102249145508,
+      "learning_rate": 6.197197197197198e-06,
+      "loss": 0.4648,
+      "step": 37990
+    },
+    {
+      "epoch": 114.11,
+      "grad_norm": 11.963196754455566,
+      "learning_rate": 6.196196196196197e-06,
+      "loss": 0.43,
+      "step": 38000
+    },
+    {
+      "epoch": 114.14,
+      "grad_norm": 18.492630004882812,
+      "learning_rate": 6.195195195195196e-06,
+      "loss": 0.4229,
+      "step": 38010
+    },
+    {
+      "epoch": 114.17,
+      "grad_norm": 15.621365547180176,
+      "learning_rate": 6.194194194194195e-06,
+      "loss": 0.3843,
+      "step": 38020
+    },
+    {
+      "epoch": 114.2,
+      "grad_norm": 15.686116218566895,
+      "learning_rate": 6.1931931931931934e-06,
+      "loss": 0.4611,
+      "step": 38030
+    },
+    {
+      "epoch": 114.23,
+      "grad_norm": 30.557775497436523,
+      "learning_rate": 6.192192192192192e-06,
+      "loss": 0.4067,
+      "step": 38040
+    },
+    {
+      "epoch": 114.26,
+      "grad_norm": 14.932382583618164,
+      "learning_rate": 6.1911911911911915e-06,
+      "loss": 0.4502,
+      "step": 38050
+    },
+    {
+      "epoch": 114.29,
+      "grad_norm": 18.036909103393555,
+      "learning_rate": 6.190190190190191e-06,
+      "loss": 0.4529,
+      "step": 38060
+    },
+    {
+      "epoch": 114.32,
+      "grad_norm": 13.089601516723633,
+      "learning_rate": 6.1891891891891895e-06,
+      "loss": 0.4282,
+      "step": 38070
+    },
+    {
+      "epoch": 114.35,
+      "grad_norm": 19.72980499267578,
+      "learning_rate": 6.188188188188189e-06,
+      "loss": 0.4971,
+      "step": 38080
+    },
+    {
+      "epoch": 114.38,
+      "grad_norm": 13.557571411132812,
+      "learning_rate": 6.187187187187188e-06,
+      "loss": 0.4204,
+      "step": 38090
+    },
+    {
+      "epoch": 114.41,
+      "grad_norm": 21.028484344482422,
+      "learning_rate": 6.186186186186187e-06,
+      "loss": 0.452,
+      "step": 38100
+    },
+    {
+      "epoch": 114.44,
+      "grad_norm": 13.464184761047363,
+      "learning_rate": 6.1851851851851856e-06,
+      "loss": 0.4497,
+      "step": 38110
+    },
+    {
+      "epoch": 114.47,
+      "grad_norm": 11.997215270996094,
+      "learning_rate": 6.184184184184184e-06,
+      "loss": 0.4725,
+      "step": 38120
+    },
+    {
+      "epoch": 114.5,
+      "grad_norm": 16.442752838134766,
+      "learning_rate": 6.183183183183184e-06,
+      "loss": 0.4289,
+      "step": 38130
+    },
+    {
+      "epoch": 114.53,
+      "grad_norm": 12.835067749023438,
+      "learning_rate": 6.182182182182182e-06,
+      "loss": 0.4457,
+      "step": 38140
+    },
+    {
+      "epoch": 114.56,
+      "grad_norm": 16.43677520751953,
+      "learning_rate": 6.181181181181182e-06,
+      "loss": 0.4444,
+      "step": 38150
+    },
+    {
+      "epoch": 114.59,
+      "grad_norm": 15.038857460021973,
+      "learning_rate": 6.180180180180181e-06,
+      "loss": 0.5077,
+      "step": 38160
+    },
+    {
+      "epoch": 114.62,
+      "grad_norm": 18.010494232177734,
+      "learning_rate": 6.17917917917918e-06,
+      "loss": 0.4595,
+      "step": 38170
+    },
+    {
+      "epoch": 114.65,
+      "grad_norm": 21.995906829833984,
+      "learning_rate": 6.178178178178179e-06,
+      "loss": 0.4201,
+      "step": 38180
+    },
+    {
+      "epoch": 114.68,
+      "grad_norm": 14.650400161743164,
+      "learning_rate": 6.177177177177177e-06,
+      "loss": 0.4551,
+      "step": 38190
+    },
+    {
+      "epoch": 114.71,
+      "grad_norm": 18.831300735473633,
+      "learning_rate": 6.176176176176176e-06,
+      "loss": 0.4509,
+      "step": 38200
+    },
+    {
+      "epoch": 114.74,
+      "grad_norm": 17.067169189453125,
+      "learning_rate": 6.175175175175176e-06,
+      "loss": 0.444,
+      "step": 38210
+    },
+    {
+      "epoch": 114.77,
+      "grad_norm": 12.900906562805176,
+      "learning_rate": 6.174174174174174e-06,
+      "loss": 0.4706,
+      "step": 38220
+    },
+    {
+      "epoch": 114.8,
+      "grad_norm": 13.785622596740723,
+      "learning_rate": 6.173173173173174e-06,
+      "loss": 0.4078,
+      "step": 38230
+    },
+    {
+      "epoch": 114.83,
+      "grad_norm": 21.62495231628418,
+      "learning_rate": 6.172172172172172e-06,
+      "loss": 0.4229,
+      "step": 38240
+    },
+    {
+      "epoch": 114.86,
+      "grad_norm": 10.005603790283203,
+      "learning_rate": 6.171171171171172e-06,
+      "loss": 0.4708,
+      "step": 38250
+    },
+    {
+      "epoch": 114.89,
+      "grad_norm": 14.358074188232422,
+      "learning_rate": 6.170170170170171e-06,
+      "loss": 0.464,
+      "step": 38260
+    },
+    {
+      "epoch": 114.92,
+      "grad_norm": 14.9757661819458,
+      "learning_rate": 6.16916916916917e-06,
+      "loss": 0.4124,
+      "step": 38270
+    },
+    {
+      "epoch": 114.95,
+      "grad_norm": 20.51958656311035,
+      "learning_rate": 6.1681681681681685e-06,
+      "loss": 0.3988,
+      "step": 38280
+    },
+    {
+      "epoch": 114.98,
+      "grad_norm": 29.6107234954834,
+      "learning_rate": 6.167167167167167e-06,
+      "loss": 0.4621,
+      "step": 38290
+    },
+    {
+      "epoch": 115.0,
+      "eval_accuracy": 0.867,
+      "eval_loss": 0.493683397769928,
+      "eval_runtime": 12.8162,
+      "eval_samples_per_second": 780.265,
+      "eval_steps_per_second": 3.121,
+      "step": 38295
+    },
+    {
+      "epoch": 115.02,
+      "grad_norm": 16.30169105529785,
+      "learning_rate": 6.1661661661661665e-06,
+      "loss": 0.3821,
+      "step": 38300
+    },
+    {
+      "epoch": 115.05,
+      "grad_norm": 12.021759986877441,
+      "learning_rate": 6.165165165165166e-06,
+      "loss": 0.4517,
+      "step": 38310
+    },
+    {
+      "epoch": 115.08,
+      "grad_norm": 15.901144981384277,
+      "learning_rate": 6.1641641641641645e-06,
+      "loss": 0.4544,
+      "step": 38320
+    },
+    {
+      "epoch": 115.11,
+      "grad_norm": 21.69875717163086,
+      "learning_rate": 6.163163163163164e-06,
+      "loss": 0.4798,
+      "step": 38330
+    },
+    {
+      "epoch": 115.14,
+      "grad_norm": 19.52529525756836,
+      "learning_rate": 6.162162162162163e-06,
+      "loss": 0.4397,
+      "step": 38340
+    },
+    {
+      "epoch": 115.17,
+      "grad_norm": 13.409546852111816,
+      "learning_rate": 6.161161161161162e-06,
+      "loss": 0.4969,
+      "step": 38350
+    },
+    {
+      "epoch": 115.2,
+      "grad_norm": 13.545701026916504,
+      "learning_rate": 6.16016016016016e-06,
+      "loss": 0.4205,
+      "step": 38360
+    },
+    {
+      "epoch": 115.23,
+      "grad_norm": 14.286934852600098,
+      "learning_rate": 6.159159159159159e-06,
+      "loss": 0.3629,
+      "step": 38370
+    },
+    {
+      "epoch": 115.26,
+      "grad_norm": 12.383866310119629,
+      "learning_rate": 6.158158158158159e-06,
+      "loss": 0.4274,
+      "step": 38380
+    },
+    {
+      "epoch": 115.29,
+      "grad_norm": 15.41568660736084,
+      "learning_rate": 6.157157157157157e-06,
+      "loss": 0.3909,
+      "step": 38390
+    },
+    {
+      "epoch": 115.32,
+      "grad_norm": 14.392060279846191,
+      "learning_rate": 6.156156156156157e-06,
+      "loss": 0.4831,
+      "step": 38400
+    },
+    {
+      "epoch": 115.35,
+      "grad_norm": 13.975421905517578,
+      "learning_rate": 6.155155155155156e-06,
+      "loss": 0.4578,
+      "step": 38410
+    },
+    {
+      "epoch": 115.38,
+      "grad_norm": 18.5854549407959,
+      "learning_rate": 6.154154154154155e-06,
+      "loss": 0.4346,
+      "step": 38420
+    },
+    {
+      "epoch": 115.41,
+      "grad_norm": 19.04314422607422,
+      "learning_rate": 6.153153153153154e-06,
+      "loss": 0.4548,
+      "step": 38430
+    },
+    {
+      "epoch": 115.44,
+      "grad_norm": 14.406473159790039,
+      "learning_rate": 6.152152152152152e-06,
+      "loss": 0.3998,
+      "step": 38440
+    },
+    {
+      "epoch": 115.47,
+      "grad_norm": 16.06531524658203,
+      "learning_rate": 6.151151151151151e-06,
+      "loss": 0.4201,
+      "step": 38450
+    },
+    {
+      "epoch": 115.5,
+      "grad_norm": 14.07009220123291,
+      "learning_rate": 6.150150150150151e-06,
+      "loss": 0.4566,
+      "step": 38460
+    },
+    {
+      "epoch": 115.53,
+      "grad_norm": 15.940716743469238,
+      "learning_rate": 6.149149149149149e-06,
+      "loss": 0.4401,
+      "step": 38470
+    },
+    {
+      "epoch": 115.56,
+      "grad_norm": 12.023195266723633,
+      "learning_rate": 6.148148148148149e-06,
+      "loss": 0.4131,
+      "step": 38480
+    },
+    {
+      "epoch": 115.59,
+      "grad_norm": 16.543386459350586,
+      "learning_rate": 6.147147147147147e-06,
+      "loss": 0.4954,
+      "step": 38490
+    },
+    {
+      "epoch": 115.62,
+      "grad_norm": 20.59006118774414,
+      "learning_rate": 6.146146146146147e-06,
+      "loss": 0.4693,
+      "step": 38500
+    },
+    {
+      "epoch": 115.65,
+      "grad_norm": 19.963863372802734,
+      "learning_rate": 6.145145145145146e-06,
+      "loss": 0.4216,
+      "step": 38510
+    },
+    {
+      "epoch": 115.68,
+      "grad_norm": 14.681634902954102,
+      "learning_rate": 6.144144144144145e-06,
+      "loss": 0.4682,
+      "step": 38520
+    },
+    {
+      "epoch": 115.71,
+      "grad_norm": 19.43003273010254,
+      "learning_rate": 6.1431431431431435e-06,
+      "loss": 0.4063,
+      "step": 38530
+    },
+    {
+      "epoch": 115.74,
+      "grad_norm": 16.149311065673828,
+      "learning_rate": 6.142142142142142e-06,
+      "loss": 0.4299,
+      "step": 38540
+    },
+    {
+      "epoch": 115.77,
+      "grad_norm": 15.602410316467285,
+      "learning_rate": 6.1411411411411415e-06,
+      "loss": 0.4693,
+      "step": 38550
+    },
+    {
+      "epoch": 115.8,
+      "grad_norm": 16.268634796142578,
+      "learning_rate": 6.140140140140141e-06,
+      "loss": 0.4863,
+      "step": 38560
+    },
+    {
+      "epoch": 115.83,
+      "grad_norm": 14.703850746154785,
+      "learning_rate": 6.1391391391391396e-06,
+      "loss": 0.4954,
+      "step": 38570
+    },
+    {
+      "epoch": 115.86,
+      "grad_norm": 16.37369728088379,
+      "learning_rate": 6.138138138138139e-06,
+      "loss": 0.4307,
+      "step": 38580
+    },
+    {
+      "epoch": 115.89,
+      "grad_norm": 20.996129989624023,
+      "learning_rate": 6.137137137137138e-06,
+      "loss": 0.4051,
+      "step": 38590
+    },
+    {
+      "epoch": 115.92,
+      "grad_norm": 14.25339412689209,
+      "learning_rate": 6.136136136136137e-06,
+      "loss": 0.5021,
+      "step": 38600
+    },
+    {
+      "epoch": 115.95,
+      "grad_norm": 15.614863395690918,
+      "learning_rate": 6.135135135135135e-06,
+      "loss": 0.4535,
+      "step": 38610
+    },
+    {
+      "epoch": 115.98,
+      "grad_norm": 17.611942291259766,
+      "learning_rate": 6.134134134134134e-06,
+      "loss": 0.467,
+      "step": 38620
+    },
+    {
+      "epoch": 116.0,
+      "eval_accuracy": 0.8671,
+      "eval_loss": 0.4961470663547516,
+      "eval_runtime": 12.8489,
+      "eval_samples_per_second": 778.278,
+      "eval_steps_per_second": 3.113,
+      "step": 38628
+    },
+    {
+      "epoch": 116.01,
+      "grad_norm": 28.08994483947754,
+      "learning_rate": 6.133133133133134e-06,
+      "loss": 0.5163,
+      "step": 38630
+    },
+    {
+      "epoch": 116.04,
+      "grad_norm": 10.827581405639648,
+      "learning_rate": 6.132132132132132e-06,
+      "loss": 0.496,
+      "step": 38640
+    },
+    {
+      "epoch": 116.07,
+      "grad_norm": 11.607084274291992,
+      "learning_rate": 6.131131131131132e-06,
+      "loss": 0.4339,
+      "step": 38650
+    },
+    {
+      "epoch": 116.1,
+      "grad_norm": 15.280745506286621,
+      "learning_rate": 6.130130130130131e-06,
+      "loss": 0.4194,
+      "step": 38660
+    },
+    {
+      "epoch": 116.13,
+      "grad_norm": 23.699813842773438,
+      "learning_rate": 6.12912912912913e-06,
+      "loss": 0.4525,
+      "step": 38670
+    },
+    {
+      "epoch": 116.16,
+      "grad_norm": 22.895368576049805,
+      "learning_rate": 6.128128128128129e-06,
+      "loss": 0.4273,
+      "step": 38680
+    },
+    {
+      "epoch": 116.19,
+      "grad_norm": 20.062986373901367,
+      "learning_rate": 6.127127127127127e-06,
+      "loss": 0.4664,
+      "step": 38690
+    },
+    {
+      "epoch": 116.22,
+      "grad_norm": 14.79334545135498,
+      "learning_rate": 6.126126126126126e-06,
+      "loss": 0.4387,
+      "step": 38700
+    },
+    {
+      "epoch": 116.25,
+      "grad_norm": 13.058863639831543,
+      "learning_rate": 6.125125125125125e-06,
+      "loss": 0.4272,
+      "step": 38710
+    },
+    {
+      "epoch": 116.28,
+      "grad_norm": 14.789053916931152,
+      "learning_rate": 6.124124124124124e-06,
+      "loss": 0.4757,
+      "step": 38720
+    },
+    {
+      "epoch": 116.31,
+      "grad_norm": 22.570030212402344,
+      "learning_rate": 6.123123123123124e-06,
+      "loss": 0.4511,
+      "step": 38730
+    },
+    {
+      "epoch": 116.34,
+      "grad_norm": 13.798760414123535,
+      "learning_rate": 6.1221221221221225e-06,
+      "loss": 0.4915,
+      "step": 38740
+    },
+    {
+      "epoch": 116.37,
+      "grad_norm": 14.481497764587402,
+      "learning_rate": 6.121121121121122e-06,
+      "loss": 0.4339,
+      "step": 38750
+    },
+    {
+      "epoch": 116.4,
+      "grad_norm": 14.194169998168945,
+      "learning_rate": 6.120120120120121e-06,
+      "loss": 0.4177,
+      "step": 38760
+    },
+    {
+      "epoch": 116.43,
+      "grad_norm": 16.108949661254883,
+      "learning_rate": 6.11911911911912e-06,
+      "loss": 0.4186,
+      "step": 38770
+    },
+    {
+      "epoch": 116.46,
+      "grad_norm": 21.837345123291016,
+      "learning_rate": 6.1181181181181185e-06,
+      "loss": 0.4371,
+      "step": 38780
+    },
+    {
+      "epoch": 116.49,
+      "grad_norm": 18.033079147338867,
+      "learning_rate": 6.117117117117117e-06,
+      "loss": 0.469,
+      "step": 38790
+    },
+    {
+      "epoch": 116.52,
+      "grad_norm": 18.654630661010742,
+      "learning_rate": 6.1161161161161166e-06,
+      "loss": 0.4162,
+      "step": 38800
+    },
+    {
+      "epoch": 116.55,
+      "grad_norm": 21.456539154052734,
+      "learning_rate": 6.115115115115115e-06,
+      "loss": 0.4605,
+      "step": 38810
+    },
+    {
+      "epoch": 116.58,
+      "grad_norm": 17.881877899169922,
+      "learning_rate": 6.114114114114115e-06,
+      "loss": 0.481,
+      "step": 38820
+    },
+    {
+      "epoch": 116.61,
+      "grad_norm": 16.79380989074707,
+      "learning_rate": 6.113113113113114e-06,
+      "loss": 0.429,
+      "step": 38830
+    },
+    {
+      "epoch": 116.64,
+      "grad_norm": 18.378036499023438,
+      "learning_rate": 6.112112112112113e-06,
+      "loss": 0.4421,
+      "step": 38840
+    },
+    {
+      "epoch": 116.67,
+      "grad_norm": 16.181352615356445,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.4524,
+      "step": 38850
+    },
+    {
+      "epoch": 116.7,
+      "grad_norm": 21.082242965698242,
+      "learning_rate": 6.11011011011011e-06,
+      "loss": 0.4543,
+      "step": 38860
+    },
+    {
+      "epoch": 116.73,
+      "grad_norm": 11.735249519348145,
+      "learning_rate": 6.109109109109109e-06,
+      "loss": 0.3403,
+      "step": 38870
+    },
+    {
+      "epoch": 116.76,
+      "grad_norm": 19.57173728942871,
+      "learning_rate": 6.108108108108109e-06,
+      "loss": 0.3994,
+      "step": 38880
+    },
+    {
+      "epoch": 116.79,
+      "grad_norm": 16.919519424438477,
+      "learning_rate": 6.107107107107107e-06,
+      "loss": 0.4644,
+      "step": 38890
+    },
+    {
+      "epoch": 116.82,
+      "grad_norm": 16.03952980041504,
+      "learning_rate": 6.106106106106107e-06,
+      "loss": 0.4278,
+      "step": 38900
+    },
+    {
+      "epoch": 116.85,
+      "grad_norm": 11.172567367553711,
+      "learning_rate": 6.105105105105106e-06,
+      "loss": 0.4118,
+      "step": 38910
+    },
+    {
+      "epoch": 116.88,
+      "grad_norm": 10.422479629516602,
+      "learning_rate": 6.104104104104105e-06,
+      "loss": 0.4049,
+      "step": 38920
+    },
+    {
+      "epoch": 116.91,
+      "grad_norm": 15.837510108947754,
+      "learning_rate": 6.103103103103104e-06,
+      "loss": 0.468,
+      "step": 38930
+    },
+    {
+      "epoch": 116.94,
+      "grad_norm": 12.611438751220703,
+      "learning_rate": 6.102102102102102e-06,
+      "loss": 0.4347,
+      "step": 38940
+    },
+    {
+      "epoch": 116.97,
+      "grad_norm": 13.27275562286377,
+      "learning_rate": 6.101101101101101e-06,
+      "loss": 0.4395,
+      "step": 38950
+    },
+    {
+      "epoch": 117.0,
+      "grad_norm": 17.27047348022461,
+      "learning_rate": 6.1001001001001e-06,
+      "loss": 0.4328,
+      "step": 38960
+    },
+    {
+      "epoch": 117.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.49869194626808167,
+      "eval_runtime": 13.0395,
+      "eval_samples_per_second": 766.899,
+      "eval_steps_per_second": 3.068,
+      "step": 38961
+    },
+    {
+      "epoch": 117.03,
+      "grad_norm": 15.880773544311523,
+      "learning_rate": 6.0990990990990995e-06,
+      "loss": 0.3856,
+      "step": 38970
+    },
+    {
+      "epoch": 117.06,
+      "grad_norm": 15.28426742553711,
+      "learning_rate": 6.098098098098099e-06,
+      "loss": 0.4446,
+      "step": 38980
+    },
+    {
+      "epoch": 117.09,
+      "grad_norm": 16.554494857788086,
+      "learning_rate": 6.0970970970970975e-06,
+      "loss": 0.385,
+      "step": 38990
+    },
+    {
+      "epoch": 117.12,
+      "grad_norm": 17.56060218811035,
+      "learning_rate": 6.096096096096097e-06,
+      "loss": 0.5106,
+      "step": 39000
+    },
+    {
+      "epoch": 117.15,
+      "grad_norm": 18.475717544555664,
+      "learning_rate": 6.095095095095096e-06,
+      "loss": 0.4039,
+      "step": 39010
+    },
+    {
+      "epoch": 117.18,
+      "grad_norm": 10.235529899597168,
+      "learning_rate": 6.094094094094095e-06,
+      "loss": 0.419,
+      "step": 39020
+    },
+    {
+      "epoch": 117.21,
+      "grad_norm": 21.031347274780273,
+      "learning_rate": 6.0930930930930936e-06,
+      "loss": 0.4263,
+      "step": 39030
+    },
+    {
+      "epoch": 117.24,
+      "grad_norm": 16.857280731201172,
+      "learning_rate": 6.092092092092092e-06,
+      "loss": 0.4222,
+      "step": 39040
+    },
+    {
+      "epoch": 117.27,
+      "grad_norm": 17.912076950073242,
+      "learning_rate": 6.091091091091092e-06,
+      "loss": 0.4414,
+      "step": 39050
+    },
+    {
+      "epoch": 117.3,
+      "grad_norm": 21.047870635986328,
+      "learning_rate": 6.09009009009009e-06,
+      "loss": 0.4216,
+      "step": 39060
+    },
+    {
+      "epoch": 117.33,
+      "grad_norm": 13.466742515563965,
+      "learning_rate": 6.08908908908909e-06,
+      "loss": 0.4851,
+      "step": 39070
+    },
+    {
+      "epoch": 117.36,
+      "grad_norm": 18.22876739501953,
+      "learning_rate": 6.088088088088089e-06,
+      "loss": 0.466,
+      "step": 39080
+    },
+    {
+      "epoch": 117.39,
+      "grad_norm": 14.583304405212402,
+      "learning_rate": 6.087087087087088e-06,
+      "loss": 0.4338,
+      "step": 39090
+    },
+    {
+      "epoch": 117.42,
+      "grad_norm": 16.644485473632812,
+      "learning_rate": 6.086086086086087e-06,
+      "loss": 0.427,
+      "step": 39100
+    },
+    {
+      "epoch": 117.45,
+      "grad_norm": 17.641815185546875,
+      "learning_rate": 6.085085085085085e-06,
+      "loss": 0.4405,
+      "step": 39110
+    },
+    {
+      "epoch": 117.48,
+      "grad_norm": 19.21428871154785,
+      "learning_rate": 6.084084084084084e-06,
+      "loss": 0.4771,
+      "step": 39120
+    },
+    {
+      "epoch": 117.51,
+      "grad_norm": 21.12874984741211,
+      "learning_rate": 6.083083083083084e-06,
+      "loss": 0.4238,
+      "step": 39130
+    },
+    {
+      "epoch": 117.54,
+      "grad_norm": 11.589507102966309,
+      "learning_rate": 6.082082082082082e-06,
+      "loss": 0.4024,
+      "step": 39140
+    },
+    {
+      "epoch": 117.57,
+      "grad_norm": 13.799592018127441,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 0.481,
+      "step": 39150
+    },
+    {
+      "epoch": 117.6,
+      "grad_norm": 12.024993896484375,
+      "learning_rate": 6.08008008008008e-06,
+      "loss": 0.4623,
+      "step": 39160
+    },
+    {
+      "epoch": 117.63,
+      "grad_norm": 11.261701583862305,
+      "learning_rate": 6.07907907907908e-06,
+      "loss": 0.4731,
+      "step": 39170
+    },
+    {
+      "epoch": 117.66,
+      "grad_norm": 17.883211135864258,
+      "learning_rate": 6.078078078078079e-06,
+      "loss": 0.4017,
+      "step": 39180
+    },
+    {
+      "epoch": 117.69,
+      "grad_norm": 15.288254737854004,
+      "learning_rate": 6.077077077077077e-06,
+      "loss": 0.4196,
+      "step": 39190
+    },
+    {
+      "epoch": 117.72,
+      "grad_norm": 14.531414031982422,
+      "learning_rate": 6.0760760760760765e-06,
+      "loss": 0.4444,
+      "step": 39200
+    },
+    {
+      "epoch": 117.75,
+      "grad_norm": 11.638463020324707,
+      "learning_rate": 6.075075075075075e-06,
+      "loss": 0.5004,
+      "step": 39210
+    },
+    {
+      "epoch": 117.78,
+      "grad_norm": 14.134969711303711,
+      "learning_rate": 6.0740740740740745e-06,
+      "loss": 0.4778,
+      "step": 39220
+    },
+    {
+      "epoch": 117.81,
+      "grad_norm": 22.74114418029785,
+      "learning_rate": 6.073073073073074e-06,
+      "loss": 0.434,
+      "step": 39230
+    },
+    {
+      "epoch": 117.84,
+      "grad_norm": 19.305246353149414,
+      "learning_rate": 6.0720720720720725e-06,
+      "loss": 0.4474,
+      "step": 39240
+    },
+    {
+      "epoch": 117.87,
+      "grad_norm": 14.76930046081543,
+      "learning_rate": 6.071071071071072e-06,
+      "loss": 0.4297,
+      "step": 39250
+    },
+    {
+      "epoch": 117.9,
+      "grad_norm": 21.92295265197754,
+      "learning_rate": 6.0700700700700706e-06,
+      "loss": 0.4491,
+      "step": 39260
+    },
+    {
+      "epoch": 117.93,
+      "grad_norm": 15.860072135925293,
+      "learning_rate": 6.06906906906907e-06,
+      "loss": 0.4956,
+      "step": 39270
+    },
+    {
+      "epoch": 117.96,
+      "grad_norm": 20.457061767578125,
+      "learning_rate": 6.068068068068068e-06,
+      "loss": 0.4319,
+      "step": 39280
+    },
+    {
+      "epoch": 117.99,
+      "grad_norm": 9.813947677612305,
+      "learning_rate": 6.067067067067067e-06,
+      "loss": 0.3697,
+      "step": 39290
+    },
+    {
+      "epoch": 118.0,
+      "eval_accuracy": 0.8667,
+      "eval_loss": 0.4968120753765106,
+      "eval_runtime": 12.7612,
+      "eval_samples_per_second": 783.624,
+      "eval_steps_per_second": 3.134,
+      "step": 39294
+    },
+    {
+      "epoch": 118.02,
+      "grad_norm": 19.99506950378418,
+      "learning_rate": 6.066066066066067e-06,
+      "loss": 0.4472,
+      "step": 39300
+    },
+    {
+      "epoch": 118.05,
+      "grad_norm": 12.236839294433594,
+      "learning_rate": 6.065065065065065e-06,
+      "loss": 0.4096,
+      "step": 39310
+    },
+    {
+      "epoch": 118.08,
+      "grad_norm": 19.289215087890625,
+      "learning_rate": 6.064064064064065e-06,
+      "loss": 0.4326,
+      "step": 39320
+    },
+    {
+      "epoch": 118.11,
+      "grad_norm": 13.688709259033203,
+      "learning_rate": 6.063063063063064e-06,
+      "loss": 0.4646,
+      "step": 39330
+    },
+    {
+      "epoch": 118.14,
+      "grad_norm": 20.05377769470215,
+      "learning_rate": 6.062062062062063e-06,
+      "loss": 0.4026,
+      "step": 39340
+    },
+    {
+      "epoch": 118.17,
+      "grad_norm": 13.346729278564453,
+      "learning_rate": 6.061061061061062e-06,
+      "loss": 0.3809,
+      "step": 39350
+    },
+    {
+      "epoch": 118.2,
+      "grad_norm": 20.537240982055664,
+      "learning_rate": 6.06006006006006e-06,
+      "loss": 0.4742,
+      "step": 39360
+    },
+    {
+      "epoch": 118.23,
+      "grad_norm": 12.952225685119629,
+      "learning_rate": 6.059059059059059e-06,
+      "loss": 0.4566,
+      "step": 39370
+    },
+    {
+      "epoch": 118.26,
+      "grad_norm": 11.028264999389648,
+      "learning_rate": 6.058058058058058e-06,
+      "loss": 0.4019,
+      "step": 39380
+    },
+    {
+      "epoch": 118.29,
+      "grad_norm": 15.54875373840332,
+      "learning_rate": 6.057057057057057e-06,
+      "loss": 0.4434,
+      "step": 39390
+    },
+    {
+      "epoch": 118.32,
+      "grad_norm": 16.326082229614258,
+      "learning_rate": 6.056056056056057e-06,
+      "loss": 0.4655,
+      "step": 39400
+    },
+    {
+      "epoch": 118.35,
+      "grad_norm": 17.78352165222168,
+      "learning_rate": 6.055055055055055e-06,
+      "loss": 0.4211,
+      "step": 39410
+    },
+    {
+      "epoch": 118.38,
+      "grad_norm": 17.679718017578125,
+      "learning_rate": 6.054054054054055e-06,
+      "loss": 0.4438,
+      "step": 39420
+    },
+    {
+      "epoch": 118.41,
+      "grad_norm": 14.511039733886719,
+      "learning_rate": 6.053053053053054e-06,
+      "loss": 0.4304,
+      "step": 39430
+    },
+    {
+      "epoch": 118.44,
+      "grad_norm": 17.333757400512695,
+      "learning_rate": 6.052052052052052e-06,
+      "loss": 0.4259,
+      "step": 39440
+    },
+    {
+      "epoch": 118.47,
+      "grad_norm": 13.99795150756836,
+      "learning_rate": 6.0510510510510515e-06,
+      "loss": 0.4086,
+      "step": 39450
+    },
+    {
+      "epoch": 118.5,
+      "grad_norm": 16.08959197998047,
+      "learning_rate": 6.05005005005005e-06,
+      "loss": 0.441,
+      "step": 39460
+    },
+    {
+      "epoch": 118.53,
+      "grad_norm": 20.08635902404785,
+      "learning_rate": 6.0490490490490495e-06,
+      "loss": 0.4851,
+      "step": 39470
+    },
+    {
+      "epoch": 118.56,
+      "grad_norm": 23.677104949951172,
+      "learning_rate": 6.048048048048049e-06,
+      "loss": 0.4752,
+      "step": 39480
+    },
+    {
+      "epoch": 118.59,
+      "grad_norm": 14.720694541931152,
+      "learning_rate": 6.0470470470470476e-06,
+      "loss": 0.4248,
+      "step": 39490
+    },
+    {
+      "epoch": 118.62,
+      "grad_norm": 15.068927764892578,
+      "learning_rate": 6.046046046046047e-06,
+      "loss": 0.4493,
+      "step": 39500
+    },
+    {
+      "epoch": 118.65,
+      "grad_norm": 14.856972694396973,
+      "learning_rate": 6.045045045045046e-06,
+      "loss": 0.4199,
+      "step": 39510
+    },
+    {
+      "epoch": 118.68,
+      "grad_norm": 12.31198787689209,
+      "learning_rate": 6.044044044044044e-06,
+      "loss": 0.3943,
+      "step": 39520
+    },
+    {
+      "epoch": 118.71,
+      "grad_norm": 13.837306022644043,
+      "learning_rate": 6.043043043043043e-06,
+      "loss": 0.4222,
+      "step": 39530
+    },
+    {
+      "epoch": 118.74,
+      "grad_norm": 35.37378692626953,
+      "learning_rate": 6.042042042042042e-06,
+      "loss": 0.4578,
+      "step": 39540
+    },
+    {
+      "epoch": 118.77,
+      "grad_norm": 26.748138427734375,
+      "learning_rate": 6.041041041041042e-06,
+      "loss": 0.4236,
+      "step": 39550
+    },
+    {
+      "epoch": 118.8,
+      "grad_norm": 19.289941787719727,
+      "learning_rate": 6.04004004004004e-06,
+      "loss": 0.4479,
+      "step": 39560
+    },
+    {
+      "epoch": 118.83,
+      "grad_norm": 10.435075759887695,
+      "learning_rate": 6.03903903903904e-06,
+      "loss": 0.4122,
+      "step": 39570
+    },
+    {
+      "epoch": 118.86,
+      "grad_norm": 15.282739639282227,
+      "learning_rate": 6.038038038038039e-06,
+      "loss": 0.4668,
+      "step": 39580
+    },
+    {
+      "epoch": 118.89,
+      "grad_norm": 10.721346855163574,
+      "learning_rate": 6.037037037037038e-06,
+      "loss": 0.4465,
+      "step": 39590
+    },
+    {
+      "epoch": 118.92,
+      "grad_norm": 13.03950309753418,
+      "learning_rate": 6.036036036036037e-06,
+      "loss": 0.4186,
+      "step": 39600
+    },
+    {
+      "epoch": 118.95,
+      "grad_norm": 13.304206848144531,
+      "learning_rate": 6.035035035035035e-06,
+      "loss": 0.4292,
+      "step": 39610
+    },
+    {
+      "epoch": 118.98,
+      "grad_norm": 17.94167137145996,
+      "learning_rate": 6.034034034034034e-06,
+      "loss": 0.4668,
+      "step": 39620
+    },
+    {
+      "epoch": 119.0,
+      "eval_accuracy": 0.8655,
+      "eval_loss": 0.501977264881134,
+      "eval_runtime": 12.75,
+      "eval_samples_per_second": 784.314,
+      "eval_steps_per_second": 3.137,
+      "step": 39627
+    },
+    {
+      "epoch": 119.01,
+      "grad_norm": 10.816543579101562,
+      "learning_rate": 6.033033033033033e-06,
+      "loss": 0.4505,
+      "step": 39630
+    },
+    {
+      "epoch": 119.04,
+      "grad_norm": 15.624786376953125,
+      "learning_rate": 6.032032032032032e-06,
+      "loss": 0.4645,
+      "step": 39640
+    },
+    {
+      "epoch": 119.07,
+      "grad_norm": 13.010736465454102,
+      "learning_rate": 6.031031031031032e-06,
+      "loss": 0.4425,
+      "step": 39650
+    },
+    {
+      "epoch": 119.1,
+      "grad_norm": 13.95332145690918,
+      "learning_rate": 6.0300300300300304e-06,
+      "loss": 0.4103,
+      "step": 39660
+    },
+    {
+      "epoch": 119.13,
+      "grad_norm": 11.104092597961426,
+      "learning_rate": 6.02902902902903e-06,
+      "loss": 0.4011,
+      "step": 39670
+    },
+    {
+      "epoch": 119.16,
+      "grad_norm": 26.681419372558594,
+      "learning_rate": 6.028028028028029e-06,
+      "loss": 0.5171,
+      "step": 39680
+    },
+    {
+      "epoch": 119.19,
+      "grad_norm": 15.023918151855469,
+      "learning_rate": 6.027027027027027e-06,
+      "loss": 0.4294,
+      "step": 39690
+    },
+    {
+      "epoch": 119.22,
+      "grad_norm": 19.98373031616211,
+      "learning_rate": 6.0260260260260265e-06,
+      "loss": 0.4402,
+      "step": 39700
+    },
+    {
+      "epoch": 119.25,
+      "grad_norm": 24.974956512451172,
+      "learning_rate": 6.025025025025025e-06,
+      "loss": 0.4579,
+      "step": 39710
+    },
+    {
+      "epoch": 119.28,
+      "grad_norm": 15.918089866638184,
+      "learning_rate": 6.0240240240240246e-06,
+      "loss": 0.4522,
+      "step": 39720
+    },
+    {
+      "epoch": 119.31,
+      "grad_norm": 11.031757354736328,
+      "learning_rate": 6.023023023023023e-06,
+      "loss": 0.472,
+      "step": 39730
+    },
+    {
+      "epoch": 119.34,
+      "grad_norm": 26.119871139526367,
+      "learning_rate": 6.022022022022023e-06,
+      "loss": 0.4547,
+      "step": 39740
+    },
+    {
+      "epoch": 119.37,
+      "grad_norm": 14.452841758728027,
+      "learning_rate": 6.021021021021022e-06,
+      "loss": 0.4326,
+      "step": 39750
+    },
+    {
+      "epoch": 119.4,
+      "grad_norm": 20.64385223388672,
+      "learning_rate": 6.020020020020021e-06,
+      "loss": 0.4443,
+      "step": 39760
+    },
+    {
+      "epoch": 119.43,
+      "grad_norm": 13.606422424316406,
+      "learning_rate": 6.019019019019019e-06,
+      "loss": 0.4341,
+      "step": 39770
+    },
+    {
+      "epoch": 119.46,
+      "grad_norm": 11.301891326904297,
+      "learning_rate": 6.018018018018018e-06,
+      "loss": 0.3767,
+      "step": 39780
+    },
+    {
+      "epoch": 119.49,
+      "grad_norm": 16.744728088378906,
+      "learning_rate": 6.017017017017017e-06,
+      "loss": 0.4278,
+      "step": 39790
+    },
+    {
+      "epoch": 119.52,
+      "grad_norm": 12.527533531188965,
+      "learning_rate": 6.016016016016017e-06,
+      "loss": 0.507,
+      "step": 39800
+    },
+    {
+      "epoch": 119.55,
+      "grad_norm": 10.866864204406738,
+      "learning_rate": 6.015015015015015e-06,
+      "loss": 0.4341,
+      "step": 39810
+    },
+    {
+      "epoch": 119.58,
+      "grad_norm": 15.105118751525879,
+      "learning_rate": 6.014014014014015e-06,
+      "loss": 0.3912,
+      "step": 39820
+    },
+    {
+      "epoch": 119.61,
+      "grad_norm": 17.555335998535156,
+      "learning_rate": 6.013013013013014e-06,
+      "loss": 0.3928,
+      "step": 39830
+    },
+    {
+      "epoch": 119.64,
+      "grad_norm": 14.168087005615234,
+      "learning_rate": 6.012012012012013e-06,
+      "loss": 0.4973,
+      "step": 39840
+    },
+    {
+      "epoch": 119.67,
+      "grad_norm": 10.744633674621582,
+      "learning_rate": 6.011011011011012e-06,
+      "loss": 0.4009,
+      "step": 39850
+    },
+    {
+      "epoch": 119.7,
+      "grad_norm": 13.574105262756348,
+      "learning_rate": 6.01001001001001e-06,
+      "loss": 0.4333,
+      "step": 39860
+    },
+    {
+      "epoch": 119.73,
+      "grad_norm": 23.983171463012695,
+      "learning_rate": 6.009009009009009e-06,
+      "loss": 0.4833,
+      "step": 39870
+    },
+    {
+      "epoch": 119.76,
+      "grad_norm": 21.86933708190918,
+      "learning_rate": 6.008008008008008e-06,
+      "loss": 0.4491,
+      "step": 39880
+    },
+    {
+      "epoch": 119.79,
+      "grad_norm": 15.147838592529297,
+      "learning_rate": 6.0070070070070074e-06,
+      "loss": 0.3926,
+      "step": 39890
+    },
+    {
+      "epoch": 119.82,
+      "grad_norm": 16.565044403076172,
+      "learning_rate": 6.006006006006007e-06,
+      "loss": 0.4676,
+      "step": 39900
+    },
+    {
+      "epoch": 119.85,
+      "grad_norm": 15.626169204711914,
+      "learning_rate": 6.0050050050050055e-06,
+      "loss": 0.4152,
+      "step": 39910
+    },
+    {
+      "epoch": 119.88,
+      "grad_norm": 14.402551651000977,
+      "learning_rate": 6.004004004004005e-06,
+      "loss": 0.4567,
+      "step": 39920
+    },
+    {
+      "epoch": 119.91,
+      "grad_norm": 21.047344207763672,
+      "learning_rate": 6.003003003003004e-06,
+      "loss": 0.4454,
+      "step": 39930
+    },
+    {
+      "epoch": 119.94,
+      "grad_norm": 16.904510498046875,
+      "learning_rate": 6.002002002002002e-06,
+      "loss": 0.4177,
+      "step": 39940
+    },
+    {
+      "epoch": 119.97,
+      "grad_norm": 14.06369400024414,
+      "learning_rate": 6.0010010010010016e-06,
+      "loss": 0.4451,
+      "step": 39950
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 42.28311538696289,
+      "learning_rate": 6e-06,
+      "loss": 0.4095,
+      "step": 39960
+    },
+    {
+      "epoch": 120.0,
+      "eval_accuracy": 0.8674,
+      "eval_loss": 0.4991842806339264,
+      "eval_runtime": 12.7174,
+      "eval_samples_per_second": 786.323,
+      "eval_steps_per_second": 3.145,
+      "step": 39960
+    },
+    {
+      "epoch": 120.03,
+      "grad_norm": 13.57719898223877,
+      "learning_rate": 5.998998998999e-06,
+      "loss": 0.4239,
+      "step": 39970
+    },
+    {
+      "epoch": 120.06,
+      "grad_norm": 16.392005920410156,
+      "learning_rate": 5.997997997997998e-06,
+      "loss": 0.4885,
+      "step": 39980
+    },
+    {
+      "epoch": 120.09,
+      "grad_norm": 28.7755069732666,
+      "learning_rate": 5.996996996996998e-06,
+      "loss": 0.407,
+      "step": 39990
+    },
+    {
+      "epoch": 120.12,
+      "grad_norm": 11.518389701843262,
+      "learning_rate": 5.995995995995997e-06,
+      "loss": 0.4107,
+      "step": 40000
+    },
+    {
+      "epoch": 120.15,
+      "grad_norm": 12.662740707397461,
+      "learning_rate": 5.994994994994996e-06,
+      "loss": 0.4231,
+      "step": 40010
+    },
+    {
+      "epoch": 120.18,
+      "grad_norm": 15.818950653076172,
+      "learning_rate": 5.993993993993994e-06,
+      "loss": 0.4506,
+      "step": 40020
+    },
+    {
+      "epoch": 120.21,
+      "grad_norm": 10.51655387878418,
+      "learning_rate": 5.992992992992993e-06,
+      "loss": 0.3837,
+      "step": 40030
+    },
+    {
+      "epoch": 120.24,
+      "grad_norm": 11.199315071105957,
+      "learning_rate": 5.991991991991992e-06,
+      "loss": 0.4328,
+      "step": 40040
+    },
+    {
+      "epoch": 120.27,
+      "grad_norm": 15.689919471740723,
+      "learning_rate": 5.990990990990992e-06,
+      "loss": 0.418,
+      "step": 40050
+    },
+    {
+      "epoch": 120.3,
+      "grad_norm": 17.732440948486328,
+      "learning_rate": 5.98998998998999e-06,
+      "loss": 0.4306,
+      "step": 40060
+    },
+    {
+      "epoch": 120.33,
+      "grad_norm": 16.205232620239258,
+      "learning_rate": 5.98898898898899e-06,
+      "loss": 0.4093,
+      "step": 40070
+    },
+    {
+      "epoch": 120.36,
+      "grad_norm": 16.3626766204834,
+      "learning_rate": 5.987987987987988e-06,
+      "loss": 0.4545,
+      "step": 40080
+    },
+    {
+      "epoch": 120.39,
+      "grad_norm": 15.827106475830078,
+      "learning_rate": 5.986986986986988e-06,
+      "loss": 0.4305,
+      "step": 40090
+    },
+    {
+      "epoch": 120.42,
+      "grad_norm": 13.199630737304688,
+      "learning_rate": 5.985985985985987e-06,
+      "loss": 0.3609,
+      "step": 40100
+    },
+    {
+      "epoch": 120.45,
+      "grad_norm": 16.263263702392578,
+      "learning_rate": 5.984984984984985e-06,
+      "loss": 0.4381,
+      "step": 40110
+    },
+    {
+      "epoch": 120.48,
+      "grad_norm": 11.020049095153809,
+      "learning_rate": 5.9839839839839844e-06,
+      "loss": 0.4292,
+      "step": 40120
+    },
+    {
+      "epoch": 120.51,
+      "grad_norm": 14.504938125610352,
+      "learning_rate": 5.982982982982983e-06,
+      "loss": 0.3528,
+      "step": 40130
+    },
+    {
+      "epoch": 120.54,
+      "grad_norm": 18.861125946044922,
+      "learning_rate": 5.9819819819819825e-06,
+      "loss": 0.4781,
+      "step": 40140
+    },
+    {
+      "epoch": 120.57,
+      "grad_norm": 19.6745662689209,
+      "learning_rate": 5.980980980980982e-06,
+      "loss": 0.4042,
+      "step": 40150
+    },
+    {
+      "epoch": 120.6,
+      "grad_norm": 15.151509284973145,
+      "learning_rate": 5.9799799799799805e-06,
+      "loss": 0.3842,
+      "step": 40160
+    },
+    {
+      "epoch": 120.63,
+      "grad_norm": 13.083545684814453,
+      "learning_rate": 5.97897897897898e-06,
+      "loss": 0.4762,
+      "step": 40170
+    },
+    {
+      "epoch": 120.66,
+      "grad_norm": 17.057626724243164,
+      "learning_rate": 5.9779779779779786e-06,
+      "loss": 0.412,
+      "step": 40180
+    },
+    {
+      "epoch": 120.69,
+      "grad_norm": 16.71431541442871,
+      "learning_rate": 5.976976976976977e-06,
+      "loss": 0.4381,
+      "step": 40190
+    },
+    {
+      "epoch": 120.72,
+      "grad_norm": 20.96782112121582,
+      "learning_rate": 5.975975975975976e-06,
+      "loss": 0.4277,
+      "step": 40200
+    },
+    {
+      "epoch": 120.75,
+      "grad_norm": 17.344873428344727,
+      "learning_rate": 5.974974974974975e-06,
+      "loss": 0.4084,
+      "step": 40210
+    },
+    {
+      "epoch": 120.78,
+      "grad_norm": 22.057979583740234,
+      "learning_rate": 5.973973973973975e-06,
+      "loss": 0.4485,
+      "step": 40220
+    },
+    {
+      "epoch": 120.81,
+      "grad_norm": 22.5100154876709,
+      "learning_rate": 5.972972972972973e-06,
+      "loss": 0.4568,
+      "step": 40230
+    },
+    {
+      "epoch": 120.84,
+      "grad_norm": 15.097611427307129,
+      "learning_rate": 5.971971971971973e-06,
+      "loss": 0.4434,
+      "step": 40240
+    },
+    {
+      "epoch": 120.87,
+      "grad_norm": 12.589529991149902,
+      "learning_rate": 5.970970970970972e-06,
+      "loss": 0.4138,
+      "step": 40250
+    },
+    {
+      "epoch": 120.9,
+      "grad_norm": 11.903385162353516,
+      "learning_rate": 5.969969969969971e-06,
+      "loss": 0.4671,
+      "step": 40260
+    },
+    {
+      "epoch": 120.93,
+      "grad_norm": 20.247034072875977,
+      "learning_rate": 5.968968968968969e-06,
+      "loss": 0.4289,
+      "step": 40270
+    },
+    {
+      "epoch": 120.96,
+      "grad_norm": 13.298256874084473,
+      "learning_rate": 5.967967967967968e-06,
+      "loss": 0.4543,
+      "step": 40280
+    },
+    {
+      "epoch": 120.99,
+      "grad_norm": 10.65036392211914,
+      "learning_rate": 5.966966966966967e-06,
+      "loss": 0.4019,
+      "step": 40290
+    },
+    {
+      "epoch": 121.0,
+      "eval_accuracy": 0.864,
+      "eval_loss": 0.5087580680847168,
+      "eval_runtime": 12.9476,
+      "eval_samples_per_second": 772.345,
+      "eval_steps_per_second": 3.089,
+      "step": 40293
+    },
+    {
+      "epoch": 121.02,
+      "grad_norm": 18.9305477142334,
+      "learning_rate": 5.965965965965966e-06,
+      "loss": 0.3924,
+      "step": 40300
+    },
+    {
+      "epoch": 121.05,
+      "grad_norm": 12.201143264770508,
+      "learning_rate": 5.964964964964965e-06,
+      "loss": 0.3956,
+      "step": 40310
+    },
+    {
+      "epoch": 121.08,
+      "grad_norm": 13.443695068359375,
+      "learning_rate": 5.963963963963965e-06,
+      "loss": 0.4443,
+      "step": 40320
+    },
+    {
+      "epoch": 121.11,
+      "grad_norm": 19.255640029907227,
+      "learning_rate": 5.962962962962963e-06,
+      "loss": 0.414,
+      "step": 40330
+    },
+    {
+      "epoch": 121.14,
+      "grad_norm": 20.250165939331055,
+      "learning_rate": 5.961961961961963e-06,
+      "loss": 0.3933,
+      "step": 40340
+    },
+    {
+      "epoch": 121.17,
+      "grad_norm": 21.177946090698242,
+      "learning_rate": 5.960960960960962e-06,
+      "loss": 0.441,
+      "step": 40350
+    },
+    {
+      "epoch": 121.2,
+      "grad_norm": 17.265300750732422,
+      "learning_rate": 5.95995995995996e-06,
+      "loss": 0.41,
+      "step": 40360
+    },
+    {
+      "epoch": 121.23,
+      "grad_norm": 14.722373008728027,
+      "learning_rate": 5.9589589589589595e-06,
+      "loss": 0.4158,
+      "step": 40370
+    },
+    {
+      "epoch": 121.26,
+      "grad_norm": 13.289302825927734,
+      "learning_rate": 5.957957957957958e-06,
+      "loss": 0.4711,
+      "step": 40380
+    },
+    {
+      "epoch": 121.29,
+      "grad_norm": 15.842864990234375,
+      "learning_rate": 5.9569569569569575e-06,
+      "loss": 0.4071,
+      "step": 40390
+    },
+    {
+      "epoch": 121.32,
+      "grad_norm": 24.064443588256836,
+      "learning_rate": 5.955955955955957e-06,
+      "loss": 0.4299,
+      "step": 40400
+    },
+    {
+      "epoch": 121.35,
+      "grad_norm": 21.846620559692383,
+      "learning_rate": 5.9549549549549556e-06,
+      "loss": 0.4798,
+      "step": 40410
+    },
+    {
+      "epoch": 121.38,
+      "grad_norm": 19.170888900756836,
+      "learning_rate": 5.953953953953955e-06,
+      "loss": 0.4194,
+      "step": 40420
+    },
+    {
+      "epoch": 121.41,
+      "grad_norm": 18.096439361572266,
+      "learning_rate": 5.952952952952954e-06,
+      "loss": 0.4393,
+      "step": 40430
+    },
+    {
+      "epoch": 121.44,
+      "grad_norm": 13.390934944152832,
+      "learning_rate": 5.951951951951952e-06,
+      "loss": 0.4243,
+      "step": 40440
+    },
+    {
+      "epoch": 121.47,
+      "grad_norm": 20.65899085998535,
+      "learning_rate": 5.950950950950951e-06,
+      "loss": 0.4455,
+      "step": 40450
+    },
+    {
+      "epoch": 121.5,
+      "grad_norm": 24.45879364013672,
+      "learning_rate": 5.94994994994995e-06,
+      "loss": 0.4715,
+      "step": 40460
+    },
+    {
+      "epoch": 121.53,
+      "grad_norm": 19.623828887939453,
+      "learning_rate": 5.94894894894895e-06,
+      "loss": 0.4435,
+      "step": 40470
+    },
+    {
+      "epoch": 121.56,
+      "grad_norm": 20.96348762512207,
+      "learning_rate": 5.947947947947948e-06,
+      "loss": 0.4573,
+      "step": 40480
+    },
+    {
+      "epoch": 121.59,
+      "grad_norm": 21.043962478637695,
+      "learning_rate": 5.946946946946948e-06,
+      "loss": 0.4695,
+      "step": 40490
+    },
+    {
+      "epoch": 121.62,
+      "grad_norm": 18.68714141845703,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.4008,
+      "step": 40500
+    },
+    {
+      "epoch": 121.65,
+      "grad_norm": 14.821083068847656,
+      "learning_rate": 5.944944944944946e-06,
+      "loss": 0.4858,
+      "step": 40510
+    },
+    {
+      "epoch": 121.68,
+      "grad_norm": 13.708003044128418,
+      "learning_rate": 5.943943943943944e-06,
+      "loss": 0.398,
+      "step": 40520
+    },
+    {
+      "epoch": 121.71,
+      "grad_norm": 16.9017333984375,
+      "learning_rate": 5.942942942942943e-06,
+      "loss": 0.4632,
+      "step": 40530
+    },
+    {
+      "epoch": 121.74,
+      "grad_norm": 15.976155281066895,
+      "learning_rate": 5.941941941941942e-06,
+      "loss": 0.437,
+      "step": 40540
+    },
+    {
+      "epoch": 121.77,
+      "grad_norm": 19.469266891479492,
+      "learning_rate": 5.940940940940941e-06,
+      "loss": 0.4217,
+      "step": 40550
+    },
+    {
+      "epoch": 121.8,
+      "grad_norm": 12.34931755065918,
+      "learning_rate": 5.93993993993994e-06,
+      "loss": 0.4125,
+      "step": 40560
+    },
+    {
+      "epoch": 121.83,
+      "grad_norm": 14.694778442382812,
+      "learning_rate": 5.93893893893894e-06,
+      "loss": 0.4351,
+      "step": 40570
+    },
+    {
+      "epoch": 121.86,
+      "grad_norm": 18.229158401489258,
+      "learning_rate": 5.9379379379379384e-06,
+      "loss": 0.428,
+      "step": 40580
+    },
+    {
+      "epoch": 121.89,
+      "grad_norm": 11.491618156433105,
+      "learning_rate": 5.936936936936938e-06,
+      "loss": 0.3961,
+      "step": 40590
+    },
+    {
+      "epoch": 121.92,
+      "grad_norm": 11.39375114440918,
+      "learning_rate": 5.935935935935936e-06,
+      "loss": 0.4331,
+      "step": 40600
+    },
+    {
+      "epoch": 121.95,
+      "grad_norm": 21.080839157104492,
+      "learning_rate": 5.934934934934935e-06,
+      "loss": 0.3879,
+      "step": 40610
+    },
+    {
+      "epoch": 121.98,
+      "grad_norm": 19.955135345458984,
+      "learning_rate": 5.9339339339339345e-06,
+      "loss": 0.3627,
+      "step": 40620
+    },
+    {
+      "epoch": 122.0,
+      "eval_accuracy": 0.8658,
+      "eval_loss": 0.5078180432319641,
+      "eval_runtime": 12.69,
+      "eval_samples_per_second": 788.024,
+      "eval_steps_per_second": 3.152,
+      "step": 40626
+    },
+    {
+      "epoch": 122.01,
+      "grad_norm": 14.609880447387695,
+      "learning_rate": 5.932932932932933e-06,
+      "loss": 0.4169,
+      "step": 40630
+    },
+    {
+      "epoch": 122.04,
+      "grad_norm": 21.106128692626953,
+      "learning_rate": 5.9319319319319326e-06,
+      "loss": 0.4713,
+      "step": 40640
+    },
+    {
+      "epoch": 122.07,
+      "grad_norm": 17.360130310058594,
+      "learning_rate": 5.930930930930931e-06,
+      "loss": 0.4211,
+      "step": 40650
+    },
+    {
+      "epoch": 122.1,
+      "grad_norm": 16.314926147460938,
+      "learning_rate": 5.929929929929931e-06,
+      "loss": 0.4147,
+      "step": 40660
+    },
+    {
+      "epoch": 122.13,
+      "grad_norm": 17.099939346313477,
+      "learning_rate": 5.92892892892893e-06,
+      "loss": 0.4107,
+      "step": 40670
+    },
+    {
+      "epoch": 122.16,
+      "grad_norm": 13.367658615112305,
+      "learning_rate": 5.927927927927929e-06,
+      "loss": 0.3484,
+      "step": 40680
+    },
+    {
+      "epoch": 122.19,
+      "grad_norm": 18.690895080566406,
+      "learning_rate": 5.926926926926927e-06,
+      "loss": 0.4422,
+      "step": 40690
+    },
+    {
+      "epoch": 122.22,
+      "grad_norm": 14.37269401550293,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.476,
+      "step": 40700
+    },
+    {
+      "epoch": 122.25,
+      "grad_norm": 14.623730659484863,
+      "learning_rate": 5.924924924924925e-06,
+      "loss": 0.4543,
+      "step": 40710
+    },
+    {
+      "epoch": 122.28,
+      "grad_norm": 23.393856048583984,
+      "learning_rate": 5.923923923923925e-06,
+      "loss": 0.4683,
+      "step": 40720
+    },
+    {
+      "epoch": 122.31,
+      "grad_norm": 12.031192779541016,
+      "learning_rate": 5.922922922922923e-06,
+      "loss": 0.4937,
+      "step": 40730
+    },
+    {
+      "epoch": 122.34,
+      "grad_norm": 19.736835479736328,
+      "learning_rate": 5.921921921921923e-06,
+      "loss": 0.4443,
+      "step": 40740
+    },
+    {
+      "epoch": 122.37,
+      "grad_norm": 16.38190269470215,
+      "learning_rate": 5.920920920920921e-06,
+      "loss": 0.4812,
+      "step": 40750
+    },
+    {
+      "epoch": 122.4,
+      "grad_norm": 13.586596488952637,
+      "learning_rate": 5.919919919919921e-06,
+      "loss": 0.3974,
+      "step": 40760
+    },
+    {
+      "epoch": 122.43,
+      "grad_norm": 24.165842056274414,
+      "learning_rate": 5.9189189189189185e-06,
+      "loss": 0.3881,
+      "step": 40770
+    },
+    {
+      "epoch": 122.46,
+      "grad_norm": 11.57736873626709,
+      "learning_rate": 5.917917917917918e-06,
+      "loss": 0.4337,
+      "step": 40780
+    },
+    {
+      "epoch": 122.49,
+      "grad_norm": 11.99826717376709,
+      "learning_rate": 5.916916916916917e-06,
+      "loss": 0.4194,
+      "step": 40790
+    },
+    {
+      "epoch": 122.52,
+      "grad_norm": 14.991349220275879,
+      "learning_rate": 5.915915915915916e-06,
+      "loss": 0.4267,
+      "step": 40800
+    },
+    {
+      "epoch": 122.55,
+      "grad_norm": 14.318185806274414,
+      "learning_rate": 5.9149149149149154e-06,
+      "loss": 0.4652,
+      "step": 40810
+    },
+    {
+      "epoch": 122.58,
+      "grad_norm": 23.15039825439453,
+      "learning_rate": 5.913913913913915e-06,
+      "loss": 0.377,
+      "step": 40820
+    },
+    {
+      "epoch": 122.61,
+      "grad_norm": 16.7618408203125,
+      "learning_rate": 5.9129129129129135e-06,
+      "loss": 0.4058,
+      "step": 40830
+    },
+    {
+      "epoch": 122.64,
+      "grad_norm": 16.831361770629883,
+      "learning_rate": 5.911911911911913e-06,
+      "loss": 0.4348,
+      "step": 40840
+    },
+    {
+      "epoch": 122.67,
+      "grad_norm": 20.942672729492188,
+      "learning_rate": 5.910910910910911e-06,
+      "loss": 0.3871,
+      "step": 40850
+    },
+    {
+      "epoch": 122.7,
+      "grad_norm": 13.65446949005127,
+      "learning_rate": 5.90990990990991e-06,
+      "loss": 0.4297,
+      "step": 40860
+    },
+    {
+      "epoch": 122.73,
+      "grad_norm": 12.67642593383789,
+      "learning_rate": 5.908908908908909e-06,
+      "loss": 0.3788,
+      "step": 40870
+    },
+    {
+      "epoch": 122.76,
+      "grad_norm": 17.668954849243164,
+      "learning_rate": 5.907907907907908e-06,
+      "loss": 0.4097,
+      "step": 40880
+    },
+    {
+      "epoch": 122.79,
+      "grad_norm": 8.459508895874023,
+      "learning_rate": 5.906906906906908e-06,
+      "loss": 0.3597,
+      "step": 40890
+    },
+    {
+      "epoch": 122.82,
+      "grad_norm": 18.750904083251953,
+      "learning_rate": 5.905905905905906e-06,
+      "loss": 0.3762,
+      "step": 40900
+    },
+    {
+      "epoch": 122.85,
+      "grad_norm": 13.716808319091797,
+      "learning_rate": 5.904904904904906e-06,
+      "loss": 0.4356,
+      "step": 40910
+    },
+    {
+      "epoch": 122.88,
+      "grad_norm": 16.919458389282227,
+      "learning_rate": 5.903903903903905e-06,
+      "loss": 0.3924,
+      "step": 40920
+    },
+    {
+      "epoch": 122.91,
+      "grad_norm": 25.062725067138672,
+      "learning_rate": 5.902902902902904e-06,
+      "loss": 0.4196,
+      "step": 40930
+    },
+    {
+      "epoch": 122.94,
+      "grad_norm": 17.898441314697266,
+      "learning_rate": 5.901901901901902e-06,
+      "loss": 0.4541,
+      "step": 40940
+    },
+    {
+      "epoch": 122.97,
+      "grad_norm": 14.860621452331543,
+      "learning_rate": 5.900900900900901e-06,
+      "loss": 0.3875,
+      "step": 40950
+    },
+    {
+      "epoch": 123.0,
+      "eval_accuracy": 0.8656,
+      "eval_loss": 0.5078828930854797,
+      "eval_runtime": 12.6809,
+      "eval_samples_per_second": 788.586,
+      "eval_steps_per_second": 3.154,
+      "step": 40959
+    },
+    {
+      "epoch": 123.0,
+      "grad_norm": 26.444568634033203,
+      "learning_rate": 5.8998998998999e-06,
+      "loss": 0.4216,
+      "step": 40960
+    },
+    {
+      "epoch": 123.03,
+      "grad_norm": 15.492753982543945,
+      "learning_rate": 5.8988988988989e-06,
+      "loss": 0.3856,
+      "step": 40970
+    },
+    {
+      "epoch": 123.06,
+      "grad_norm": 17.580074310302734,
+      "learning_rate": 5.897897897897898e-06,
+      "loss": 0.45,
+      "step": 40980
+    },
+    {
+      "epoch": 123.09,
+      "grad_norm": 9.826624870300293,
+      "learning_rate": 5.896896896896898e-06,
+      "loss": 0.3904,
+      "step": 40990
+    },
+    {
+      "epoch": 123.12,
+      "grad_norm": 10.414690017700195,
+      "learning_rate": 5.895895895895896e-06,
+      "loss": 0.3948,
+      "step": 41000
+    },
+    {
+      "epoch": 123.15,
+      "grad_norm": 14.683109283447266,
+      "learning_rate": 5.894894894894896e-06,
+      "loss": 0.4082,
+      "step": 41010
+    },
+    {
+      "epoch": 123.18,
+      "grad_norm": 24.154897689819336,
+      "learning_rate": 5.8938938938938936e-06,
+      "loss": 0.3945,
+      "step": 41020
+    },
+    {
+      "epoch": 123.21,
+      "grad_norm": 15.824586868286133,
+      "learning_rate": 5.892892892892893e-06,
+      "loss": 0.4654,
+      "step": 41030
+    },
+    {
+      "epoch": 123.24,
+      "grad_norm": 13.764195442199707,
+      "learning_rate": 5.8918918918918924e-06,
+      "loss": 0.4342,
+      "step": 41040
+    },
+    {
+      "epoch": 123.27,
+      "grad_norm": 22.082311630249023,
+      "learning_rate": 5.890890890890891e-06,
+      "loss": 0.4358,
+      "step": 41050
+    },
+    {
+      "epoch": 123.3,
+      "grad_norm": 20.750450134277344,
+      "learning_rate": 5.8898898898898905e-06,
+      "loss": 0.4354,
+      "step": 41060
+    },
+    {
+      "epoch": 123.33,
+      "grad_norm": 14.417377471923828,
+      "learning_rate": 5.88888888888889e-06,
+      "loss": 0.4244,
+      "step": 41070
+    },
+    {
+      "epoch": 123.36,
+      "grad_norm": 21.26949691772461,
+      "learning_rate": 5.8878878878878885e-06,
+      "loss": 0.4326,
+      "step": 41080
+    },
+    {
+      "epoch": 123.39,
+      "grad_norm": 11.946798324584961,
+      "learning_rate": 5.886886886886888e-06,
+      "loss": 0.4909,
+      "step": 41090
+    },
+    {
+      "epoch": 123.42,
+      "grad_norm": 11.6593599319458,
+      "learning_rate": 5.885885885885886e-06,
+      "loss": 0.4099,
+      "step": 41100
+    },
+    {
+      "epoch": 123.45,
+      "grad_norm": 13.930620193481445,
+      "learning_rate": 5.884884884884885e-06,
+      "loss": 0.4375,
+      "step": 41110
+    },
+    {
+      "epoch": 123.48,
+      "grad_norm": 21.72773551940918,
+      "learning_rate": 5.883883883883884e-06,
+      "loss": 0.436,
+      "step": 41120
+    },
+    {
+      "epoch": 123.51,
+      "grad_norm": 17.99151611328125,
+      "learning_rate": 5.882882882882883e-06,
+      "loss": 0.4292,
+      "step": 41130
+    },
+    {
+      "epoch": 123.54,
+      "grad_norm": 11.573013305664062,
+      "learning_rate": 5.881881881881883e-06,
+      "loss": 0.4683,
+      "step": 41140
+    },
+    {
+      "epoch": 123.57,
+      "grad_norm": 16.794368743896484,
+      "learning_rate": 5.880880880880881e-06,
+      "loss": 0.4906,
+      "step": 41150
+    },
+    {
+      "epoch": 123.6,
+      "grad_norm": 13.283642768859863,
+      "learning_rate": 5.879879879879881e-06,
+      "loss": 0.4255,
+      "step": 41160
+    },
+    {
+      "epoch": 123.63,
+      "grad_norm": 16.809715270996094,
+      "learning_rate": 5.87887887887888e-06,
+      "loss": 0.4406,
+      "step": 41170
+    },
+    {
+      "epoch": 123.66,
+      "grad_norm": 14.091669082641602,
+      "learning_rate": 5.877877877877879e-06,
+      "loss": 0.4106,
+      "step": 41180
+    },
+    {
+      "epoch": 123.69,
+      "grad_norm": 12.341287612915039,
+      "learning_rate": 5.876876876876877e-06,
+      "loss": 0.4107,
+      "step": 41190
+    },
+    {
+      "epoch": 123.72,
+      "grad_norm": 17.921100616455078,
+      "learning_rate": 5.875875875875876e-06,
+      "loss": 0.4395,
+      "step": 41200
+    },
+    {
+      "epoch": 123.75,
+      "grad_norm": 20.59768295288086,
+      "learning_rate": 5.874874874874875e-06,
+      "loss": 0.47,
+      "step": 41210
+    },
+    {
+      "epoch": 123.78,
+      "grad_norm": 14.848880767822266,
+      "learning_rate": 5.873873873873874e-06,
+      "loss": 0.4454,
+      "step": 41220
+    },
+    {
+      "epoch": 123.81,
+      "grad_norm": 18.650638580322266,
+      "learning_rate": 5.872872872872873e-06,
+      "loss": 0.4107,
+      "step": 41230
+    },
+    {
+      "epoch": 123.84,
+      "grad_norm": 31.08252716064453,
+      "learning_rate": 5.871871871871873e-06,
+      "loss": 0.4646,
+      "step": 41240
+    },
+    {
+      "epoch": 123.87,
+      "grad_norm": 18.263717651367188,
+      "learning_rate": 5.870870870870871e-06,
+      "loss": 0.4188,
+      "step": 41250
+    },
+    {
+      "epoch": 123.9,
+      "grad_norm": 18.14675521850586,
+      "learning_rate": 5.869869869869871e-06,
+      "loss": 0.4342,
+      "step": 41260
+    },
+    {
+      "epoch": 123.93,
+      "grad_norm": 24.251392364501953,
+      "learning_rate": 5.868868868868869e-06,
+      "loss": 0.4355,
+      "step": 41270
+    },
+    {
+      "epoch": 123.96,
+      "grad_norm": 14.229436874389648,
+      "learning_rate": 5.867867867867868e-06,
+      "loss": 0.4458,
+      "step": 41280
+    },
+    {
+      "epoch": 123.99,
+      "grad_norm": 14.667645454406738,
+      "learning_rate": 5.8668668668668675e-06,
+      "loss": 0.4696,
+      "step": 41290
+    },
+    {
+      "epoch": 124.0,
+      "eval_accuracy": 0.8653,
+      "eval_loss": 0.5006229877471924,
+      "eval_runtime": 12.8019,
+      "eval_samples_per_second": 781.132,
+      "eval_steps_per_second": 3.125,
+      "step": 41292
+    },
+    {
+      "epoch": 124.02,
+      "grad_norm": 12.770699501037598,
+      "learning_rate": 5.865865865865866e-06,
+      "loss": 0.3668,
+      "step": 41300
+    },
+    {
+      "epoch": 124.05,
+      "grad_norm": 11.845354080200195,
+      "learning_rate": 5.8648648648648655e-06,
+      "loss": 0.4548,
+      "step": 41310
+    },
+    {
+      "epoch": 124.08,
+      "grad_norm": 14.77760124206543,
+      "learning_rate": 5.863863863863864e-06,
+      "loss": 0.4439,
+      "step": 41320
+    },
+    {
+      "epoch": 124.11,
+      "grad_norm": 13.230063438415527,
+      "learning_rate": 5.8628628628628635e-06,
+      "loss": 0.444,
+      "step": 41330
+    },
+    {
+      "epoch": 124.14,
+      "grad_norm": 17.53010368347168,
+      "learning_rate": 5.861861861861863e-06,
+      "loss": 0.451,
+      "step": 41340
+    },
+    {
+      "epoch": 124.17,
+      "grad_norm": 23.589305877685547,
+      "learning_rate": 5.860860860860861e-06,
+      "loss": 0.4071,
+      "step": 41350
+    },
+    {
+      "epoch": 124.2,
+      "grad_norm": 14.805013656616211,
+      "learning_rate": 5.85985985985986e-06,
+      "loss": 0.4367,
+      "step": 41360
+    },
+    {
+      "epoch": 124.23,
+      "grad_norm": 32.085365295410156,
+      "learning_rate": 5.858858858858859e-06,
+      "loss": 0.4392,
+      "step": 41370
+    },
+    {
+      "epoch": 124.26,
+      "grad_norm": 12.1990327835083,
+      "learning_rate": 5.857857857857858e-06,
+      "loss": 0.4025,
+      "step": 41380
+    },
+    {
+      "epoch": 124.29,
+      "grad_norm": 17.1855525970459,
+      "learning_rate": 5.856856856856858e-06,
+      "loss": 0.4319,
+      "step": 41390
+    },
+    {
+      "epoch": 124.32,
+      "grad_norm": 17.89725685119629,
+      "learning_rate": 5.855855855855856e-06,
+      "loss": 0.3818,
+      "step": 41400
+    },
+    {
+      "epoch": 124.35,
+      "grad_norm": 12.585066795349121,
+      "learning_rate": 5.854854854854856e-06,
+      "loss": 0.4166,
+      "step": 41410
+    },
+    {
+      "epoch": 124.38,
+      "grad_norm": 10.543429374694824,
+      "learning_rate": 5.853853853853855e-06,
+      "loss": 0.4251,
+      "step": 41420
+    },
+    {
+      "epoch": 124.41,
+      "grad_norm": 23.592565536499023,
+      "learning_rate": 5.852852852852854e-06,
+      "loss": 0.469,
+      "step": 41430
+    },
+    {
+      "epoch": 124.44,
+      "grad_norm": 15.775277137756348,
+      "learning_rate": 5.8518518518518515e-06,
+      "loss": 0.4618,
+      "step": 41440
+    },
+    {
+      "epoch": 124.47,
+      "grad_norm": 12.879891395568848,
+      "learning_rate": 5.850850850850851e-06,
+      "loss": 0.501,
+      "step": 41450
+    },
+    {
+      "epoch": 124.5,
+      "grad_norm": 17.640539169311523,
+      "learning_rate": 5.84984984984985e-06,
+      "loss": 0.381,
+      "step": 41460
+    },
+    {
+      "epoch": 124.53,
+      "grad_norm": 12.234001159667969,
+      "learning_rate": 5.848848848848849e-06,
+      "loss": 0.4168,
+      "step": 41470
+    },
+    {
+      "epoch": 124.56,
+      "grad_norm": 21.485387802124023,
+      "learning_rate": 5.847847847847848e-06,
+      "loss": 0.4414,
+      "step": 41480
+    },
+    {
+      "epoch": 124.59,
+      "grad_norm": 13.020549774169922,
+      "learning_rate": 5.846846846846848e-06,
+      "loss": 0.4499,
+      "step": 41490
+    },
+    {
+      "epoch": 124.62,
+      "grad_norm": 12.67187786102295,
+      "learning_rate": 5.8458458458458464e-06,
+      "loss": 0.4574,
+      "step": 41500
+    },
+    {
+      "epoch": 124.65,
+      "grad_norm": 19.824064254760742,
+      "learning_rate": 5.844844844844846e-06,
+      "loss": 0.457,
+      "step": 41510
+    },
+    {
+      "epoch": 124.68,
+      "grad_norm": 12.933441162109375,
+      "learning_rate": 5.843843843843844e-06,
+      "loss": 0.4208,
+      "step": 41520
+    },
+    {
+      "epoch": 124.71,
+      "grad_norm": 15.713970184326172,
+      "learning_rate": 5.842842842842843e-06,
+      "loss": 0.4384,
+      "step": 41530
+    },
+    {
+      "epoch": 124.74,
+      "grad_norm": 14.991142272949219,
+      "learning_rate": 5.8418418418418425e-06,
+      "loss": 0.4122,
+      "step": 41540
+    },
+    {
+      "epoch": 124.77,
+      "grad_norm": 13.452343940734863,
+      "learning_rate": 5.840840840840841e-06,
+      "loss": 0.4156,
+      "step": 41550
+    },
+    {
+      "epoch": 124.8,
+      "grad_norm": 11.275696754455566,
+      "learning_rate": 5.8398398398398405e-06,
+      "loss": 0.4256,
+      "step": 41560
+    },
+    {
+      "epoch": 124.83,
+      "grad_norm": 14.068809509277344,
+      "learning_rate": 5.838838838838839e-06,
+      "loss": 0.4129,
+      "step": 41570
+    },
+    {
+      "epoch": 124.86,
+      "grad_norm": 15.513654708862305,
+      "learning_rate": 5.837837837837839e-06,
+      "loss": 0.4349,
+      "step": 41580
+    },
+    {
+      "epoch": 124.89,
+      "grad_norm": 21.85318374633789,
+      "learning_rate": 5.836836836836838e-06,
+      "loss": 0.4029,
+      "step": 41590
+    },
+    {
+      "epoch": 124.92,
+      "grad_norm": 15.893622398376465,
+      "learning_rate": 5.835835835835836e-06,
+      "loss": 0.4446,
+      "step": 41600
+    },
+    {
+      "epoch": 124.95,
+      "grad_norm": 13.863004684448242,
+      "learning_rate": 5.834834834834835e-06,
+      "loss": 0.466,
+      "step": 41610
+    },
+    {
+      "epoch": 124.98,
+      "grad_norm": 15.431995391845703,
+      "learning_rate": 5.833833833833834e-06,
+      "loss": 0.4071,
+      "step": 41620
+    },
+    {
+      "epoch": 125.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5089383721351624,
+      "eval_runtime": 12.7814,
+      "eval_samples_per_second": 782.387,
+      "eval_steps_per_second": 3.13,
+      "step": 41625
+    },
+    {
+      "epoch": 125.02,
+      "grad_norm": 12.03533935546875,
+      "learning_rate": 5.832832832832833e-06,
+      "loss": 0.3322,
+      "step": 41630
+    },
+    {
+      "epoch": 125.05,
+      "grad_norm": 18.870399475097656,
+      "learning_rate": 5.831831831831833e-06,
+      "loss": 0.4014,
+      "step": 41640
+    },
+    {
+      "epoch": 125.08,
+      "grad_norm": 6.763210296630859,
+      "learning_rate": 5.830830830830831e-06,
+      "loss": 0.446,
+      "step": 41650
+    },
+    {
+      "epoch": 125.11,
+      "grad_norm": 16.18561553955078,
+      "learning_rate": 5.829829829829831e-06,
+      "loss": 0.4052,
+      "step": 41660
+    },
+    {
+      "epoch": 125.14,
+      "grad_norm": 10.347111701965332,
+      "learning_rate": 5.828828828828829e-06,
+      "loss": 0.4323,
+      "step": 41670
+    },
+    {
+      "epoch": 125.17,
+      "grad_norm": 11.702937126159668,
+      "learning_rate": 5.827827827827829e-06,
+      "loss": 0.4277,
+      "step": 41680
+    },
+    {
+      "epoch": 125.2,
+      "grad_norm": 15.218894004821777,
+      "learning_rate": 5.8268268268268265e-06,
+      "loss": 0.4263,
+      "step": 41690
+    },
+    {
+      "epoch": 125.23,
+      "grad_norm": 21.67129135131836,
+      "learning_rate": 5.825825825825826e-06,
+      "loss": 0.4574,
+      "step": 41700
+    },
+    {
+      "epoch": 125.26,
+      "grad_norm": 19.42561149597168,
+      "learning_rate": 5.824824824824825e-06,
+      "loss": 0.4371,
+      "step": 41710
+    },
+    {
+      "epoch": 125.29,
+      "grad_norm": 22.37870216369629,
+      "learning_rate": 5.823823823823824e-06,
+      "loss": 0.419,
+      "step": 41720
+    },
+    {
+      "epoch": 125.32,
+      "grad_norm": 11.0955810546875,
+      "learning_rate": 5.8228228228228234e-06,
+      "loss": 0.4296,
+      "step": 41730
+    },
+    {
+      "epoch": 125.35,
+      "grad_norm": 11.272607803344727,
+      "learning_rate": 5.821821821821823e-06,
+      "loss": 0.3961,
+      "step": 41740
+    },
+    {
+      "epoch": 125.38,
+      "grad_norm": 11.226766586303711,
+      "learning_rate": 5.8208208208208215e-06,
+      "loss": 0.3609,
+      "step": 41750
+    },
+    {
+      "epoch": 125.41,
+      "grad_norm": 14.458915710449219,
+      "learning_rate": 5.819819819819821e-06,
+      "loss": 0.398,
+      "step": 41760
+    },
+    {
+      "epoch": 125.44,
+      "grad_norm": 14.289220809936523,
+      "learning_rate": 5.818818818818819e-06,
+      "loss": 0.3898,
+      "step": 41770
+    },
+    {
+      "epoch": 125.47,
+      "grad_norm": 16.790237426757812,
+      "learning_rate": 5.817817817817818e-06,
+      "loss": 0.4608,
+      "step": 41780
+    },
+    {
+      "epoch": 125.5,
+      "grad_norm": 11.212175369262695,
+      "learning_rate": 5.816816816816817e-06,
+      "loss": 0.4143,
+      "step": 41790
+    },
+    {
+      "epoch": 125.53,
+      "grad_norm": 11.401387214660645,
+      "learning_rate": 5.815815815815816e-06,
+      "loss": 0.4288,
+      "step": 41800
+    },
+    {
+      "epoch": 125.56,
+      "grad_norm": 17.792831420898438,
+      "learning_rate": 5.814814814814816e-06,
+      "loss": 0.4799,
+      "step": 41810
+    },
+    {
+      "epoch": 125.59,
+      "grad_norm": 15.269808769226074,
+      "learning_rate": 5.813813813813814e-06,
+      "loss": 0.4031,
+      "step": 41820
+    },
+    {
+      "epoch": 125.62,
+      "grad_norm": 14.06704044342041,
+      "learning_rate": 5.812812812812814e-06,
+      "loss": 0.4501,
+      "step": 41830
+    },
+    {
+      "epoch": 125.65,
+      "grad_norm": 12.334718704223633,
+      "learning_rate": 5.811811811811813e-06,
+      "loss": 0.3594,
+      "step": 41840
+    },
+    {
+      "epoch": 125.68,
+      "grad_norm": 15.29800796508789,
+      "learning_rate": 5.810810810810811e-06,
+      "loss": 0.4452,
+      "step": 41850
+    },
+    {
+      "epoch": 125.71,
+      "grad_norm": 12.857048988342285,
+      "learning_rate": 5.80980980980981e-06,
+      "loss": 0.3801,
+      "step": 41860
+    },
+    {
+      "epoch": 125.74,
+      "grad_norm": 15.13252067565918,
+      "learning_rate": 5.808808808808809e-06,
+      "loss": 0.4526,
+      "step": 41870
+    },
+    {
+      "epoch": 125.77,
+      "grad_norm": 13.945093154907227,
+      "learning_rate": 5.807807807807808e-06,
+      "loss": 0.4133,
+      "step": 41880
+    },
+    {
+      "epoch": 125.8,
+      "grad_norm": 17.50094985961914,
+      "learning_rate": 5.806806806806808e-06,
+      "loss": 0.4406,
+      "step": 41890
+    },
+    {
+      "epoch": 125.83,
+      "grad_norm": 21.50372886657715,
+      "learning_rate": 5.805805805805806e-06,
+      "loss": 0.4532,
+      "step": 41900
+    },
+    {
+      "epoch": 125.86,
+      "grad_norm": 13.087926864624023,
+      "learning_rate": 5.804804804804806e-06,
+      "loss": 0.4123,
+      "step": 41910
+    },
+    {
+      "epoch": 125.89,
+      "grad_norm": 10.960792541503906,
+      "learning_rate": 5.803803803803804e-06,
+      "loss": 0.3967,
+      "step": 41920
+    },
+    {
+      "epoch": 125.92,
+      "grad_norm": 17.817121505737305,
+      "learning_rate": 5.802802802802803e-06,
+      "loss": 0.4692,
+      "step": 41930
+    },
+    {
+      "epoch": 125.95,
+      "grad_norm": 12.942837715148926,
+      "learning_rate": 5.8018018018018015e-06,
+      "loss": 0.4325,
+      "step": 41940
+    },
+    {
+      "epoch": 125.98,
+      "grad_norm": 15.403200149536133,
+      "learning_rate": 5.800800800800801e-06,
+      "loss": 0.4485,
+      "step": 41950
+    },
+    {
+      "epoch": 126.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.5067380666732788,
+      "eval_runtime": 12.4891,
+      "eval_samples_per_second": 800.696,
+      "eval_steps_per_second": 3.203,
+      "step": 41958
+    },
+    {
+      "epoch": 126.01,
+      "grad_norm": 13.31973934173584,
+      "learning_rate": 5.7997997997998004e-06,
+      "loss": 0.3801,
+      "step": 41960
+    },
+    {
+      "epoch": 126.04,
+      "grad_norm": 20.388015747070312,
+      "learning_rate": 5.798798798798799e-06,
+      "loss": 0.4023,
+      "step": 41970
+    },
+    {
+      "epoch": 126.07,
+      "grad_norm": 10.011873245239258,
+      "learning_rate": 5.7977977977977985e-06,
+      "loss": 0.4244,
+      "step": 41980
+    },
+    {
+      "epoch": 126.1,
+      "grad_norm": 19.137434005737305,
+      "learning_rate": 5.796796796796798e-06,
+      "loss": 0.4185,
+      "step": 41990
+    },
+    {
+      "epoch": 126.13,
+      "grad_norm": 24.39142608642578,
+      "learning_rate": 5.7957957957957965e-06,
+      "loss": 0.4102,
+      "step": 42000
+    },
+    {
+      "epoch": 126.16,
+      "grad_norm": 14.747649192810059,
+      "learning_rate": 5.794794794794796e-06,
+      "loss": 0.4425,
+      "step": 42010
+    },
+    {
+      "epoch": 126.19,
+      "grad_norm": 11.295709609985352,
+      "learning_rate": 5.793793793793794e-06,
+      "loss": 0.3792,
+      "step": 42020
+    },
+    {
+      "epoch": 126.22,
+      "grad_norm": 16.289583206176758,
+      "learning_rate": 5.792792792792793e-06,
+      "loss": 0.4025,
+      "step": 42030
+    },
+    {
+      "epoch": 126.25,
+      "grad_norm": 16.426252365112305,
+      "learning_rate": 5.791791791791792e-06,
+      "loss": 0.4723,
+      "step": 42040
+    },
+    {
+      "epoch": 126.28,
+      "grad_norm": 14.465453147888184,
+      "learning_rate": 5.790790790790791e-06,
+      "loss": 0.4551,
+      "step": 42050
+    },
+    {
+      "epoch": 126.31,
+      "grad_norm": 14.297544479370117,
+      "learning_rate": 5.789789789789791e-06,
+      "loss": 0.4675,
+      "step": 42060
+    },
+    {
+      "epoch": 126.34,
+      "grad_norm": 15.414885520935059,
+      "learning_rate": 5.788788788788789e-06,
+      "loss": 0.4351,
+      "step": 42070
+    },
+    {
+      "epoch": 126.37,
+      "grad_norm": 15.587799072265625,
+      "learning_rate": 5.787787787787789e-06,
+      "loss": 0.3786,
+      "step": 42080
+    },
+    {
+      "epoch": 126.4,
+      "grad_norm": 28.173328399658203,
+      "learning_rate": 5.786786786786788e-06,
+      "loss": 0.4668,
+      "step": 42090
+    },
+    {
+      "epoch": 126.43,
+      "grad_norm": 23.111629486083984,
+      "learning_rate": 5.785785785785786e-06,
+      "loss": 0.4371,
+      "step": 42100
+    },
+    {
+      "epoch": 126.46,
+      "grad_norm": 23.988487243652344,
+      "learning_rate": 5.784784784784785e-06,
+      "loss": 0.4467,
+      "step": 42110
+    },
+    {
+      "epoch": 126.49,
+      "grad_norm": 11.15615463256836,
+      "learning_rate": 5.783783783783784e-06,
+      "loss": 0.4135,
+      "step": 42120
+    },
+    {
+      "epoch": 126.52,
+      "grad_norm": 21.86011505126953,
+      "learning_rate": 5.782782782782783e-06,
+      "loss": 0.4377,
+      "step": 42130
+    },
+    {
+      "epoch": 126.55,
+      "grad_norm": 15.045957565307617,
+      "learning_rate": 5.781781781781782e-06,
+      "loss": 0.3544,
+      "step": 42140
+    },
+    {
+      "epoch": 126.58,
+      "grad_norm": 13.491559028625488,
+      "learning_rate": 5.780780780780781e-06,
+      "loss": 0.4329,
+      "step": 42150
+    },
+    {
+      "epoch": 126.61,
+      "grad_norm": 17.145469665527344,
+      "learning_rate": 5.779779779779781e-06,
+      "loss": 0.4175,
+      "step": 42160
+    },
+    {
+      "epoch": 126.64,
+      "grad_norm": 16.91498374938965,
+      "learning_rate": 5.778778778778779e-06,
+      "loss": 0.4611,
+      "step": 42170
+    },
+    {
+      "epoch": 126.67,
+      "grad_norm": 19.599964141845703,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 0.4617,
+      "step": 42180
+    },
+    {
+      "epoch": 126.7,
+      "grad_norm": 13.00763988494873,
+      "learning_rate": 5.776776776776777e-06,
+      "loss": 0.3888,
+      "step": 42190
+    },
+    {
+      "epoch": 126.73,
+      "grad_norm": 19.5018367767334,
+      "learning_rate": 5.775775775775776e-06,
+      "loss": 0.4604,
+      "step": 42200
+    },
+    {
+      "epoch": 126.76,
+      "grad_norm": 17.36009407043457,
+      "learning_rate": 5.7747747747747755e-06,
+      "loss": 0.4211,
+      "step": 42210
+    },
+    {
+      "epoch": 126.79,
+      "grad_norm": 14.117048263549805,
+      "learning_rate": 5.773773773773774e-06,
+      "loss": 0.3799,
+      "step": 42220
+    },
+    {
+      "epoch": 126.82,
+      "grad_norm": 15.935215950012207,
+      "learning_rate": 5.7727727727727735e-06,
+      "loss": 0.432,
+      "step": 42230
+    },
+    {
+      "epoch": 126.85,
+      "grad_norm": 15.30367660522461,
+      "learning_rate": 5.771771771771772e-06,
+      "loss": 0.4331,
+      "step": 42240
+    },
+    {
+      "epoch": 126.88,
+      "grad_norm": 14.998530387878418,
+      "learning_rate": 5.7707707707707715e-06,
+      "loss": 0.4222,
+      "step": 42250
+    },
+    {
+      "epoch": 126.91,
+      "grad_norm": 19.396326065063477,
+      "learning_rate": 5.769769769769771e-06,
+      "loss": 0.4416,
+      "step": 42260
+    },
+    {
+      "epoch": 126.94,
+      "grad_norm": 19.476715087890625,
+      "learning_rate": 5.768768768768769e-06,
+      "loss": 0.4276,
+      "step": 42270
+    },
+    {
+      "epoch": 126.97,
+      "grad_norm": 17.28748321533203,
+      "learning_rate": 5.767767767767768e-06,
+      "loss": 0.4259,
+      "step": 42280
+    },
+    {
+      "epoch": 127.0,
+      "grad_norm": 15.611334800720215,
+      "learning_rate": 5.766766766766767e-06,
+      "loss": 0.4565,
+      "step": 42290
+    },
+    {
+      "epoch": 127.0,
+      "eval_accuracy": 0.8644,
+      "eval_loss": 0.505954921245575,
+      "eval_runtime": 12.8658,
+      "eval_samples_per_second": 777.254,
+      "eval_steps_per_second": 3.109,
+      "step": 42291
+    },
+    {
+      "epoch": 127.03,
+      "grad_norm": 22.648569107055664,
+      "learning_rate": 5.765765765765766e-06,
+      "loss": 0.3814,
+      "step": 42300
+    },
+    {
+      "epoch": 127.06,
+      "grad_norm": 13.588287353515625,
+      "learning_rate": 5.764764764764766e-06,
+      "loss": 0.3971,
+      "step": 42310
+    },
+    {
+      "epoch": 127.09,
+      "grad_norm": 13.861530303955078,
+      "learning_rate": 5.763763763763764e-06,
+      "loss": 0.3769,
+      "step": 42320
+    },
+    {
+      "epoch": 127.12,
+      "grad_norm": 15.877714157104492,
+      "learning_rate": 5.762762762762764e-06,
+      "loss": 0.4588,
+      "step": 42330
+    },
+    {
+      "epoch": 127.15,
+      "grad_norm": 24.041242599487305,
+      "learning_rate": 5.761761761761763e-06,
+      "loss": 0.5328,
+      "step": 42340
+    },
+    {
+      "epoch": 127.18,
+      "grad_norm": 11.093282699584961,
+      "learning_rate": 5.760760760760761e-06,
+      "loss": 0.4105,
+      "step": 42350
+    },
+    {
+      "epoch": 127.21,
+      "grad_norm": 13.027799606323242,
+      "learning_rate": 5.7597597597597595e-06,
+      "loss": 0.4246,
+      "step": 42360
+    },
+    {
+      "epoch": 127.24,
+      "grad_norm": 12.08642578125,
+      "learning_rate": 5.758758758758759e-06,
+      "loss": 0.4282,
+      "step": 42370
+    },
+    {
+      "epoch": 127.27,
+      "grad_norm": 17.081663131713867,
+      "learning_rate": 5.757757757757758e-06,
+      "loss": 0.455,
+      "step": 42380
+    },
+    {
+      "epoch": 127.3,
+      "grad_norm": 13.423673629760742,
+      "learning_rate": 5.756756756756757e-06,
+      "loss": 0.4034,
+      "step": 42390
+    },
+    {
+      "epoch": 127.33,
+      "grad_norm": 17.521007537841797,
+      "learning_rate": 5.755755755755756e-06,
+      "loss": 0.4487,
+      "step": 42400
+    },
+    {
+      "epoch": 127.36,
+      "grad_norm": 19.942216873168945,
+      "learning_rate": 5.754754754754756e-06,
+      "loss": 0.3971,
+      "step": 42410
+    },
+    {
+      "epoch": 127.39,
+      "grad_norm": 15.989952087402344,
+      "learning_rate": 5.7537537537537544e-06,
+      "loss": 0.4012,
+      "step": 42420
+    },
+    {
+      "epoch": 127.42,
+      "grad_norm": 14.816668510437012,
+      "learning_rate": 5.752752752752753e-06,
+      "loss": 0.4699,
+      "step": 42430
+    },
+    {
+      "epoch": 127.45,
+      "grad_norm": 15.325766563415527,
+      "learning_rate": 5.751751751751752e-06,
+      "loss": 0.4309,
+      "step": 42440
+    },
+    {
+      "epoch": 127.48,
+      "grad_norm": 13.5861177444458,
+      "learning_rate": 5.750750750750751e-06,
+      "loss": 0.4013,
+      "step": 42450
+    },
+    {
+      "epoch": 127.51,
+      "grad_norm": 14.071264266967773,
+      "learning_rate": 5.7497497497497505e-06,
+      "loss": 0.4196,
+      "step": 42460
+    },
+    {
+      "epoch": 127.54,
+      "grad_norm": 11.159459114074707,
+      "learning_rate": 5.748748748748749e-06,
+      "loss": 0.4009,
+      "step": 42470
+    },
+    {
+      "epoch": 127.57,
+      "grad_norm": 13.083843231201172,
+      "learning_rate": 5.7477477477477485e-06,
+      "loss": 0.4298,
+      "step": 42480
+    },
+    {
+      "epoch": 127.6,
+      "grad_norm": 12.857629776000977,
+      "learning_rate": 5.746746746746747e-06,
+      "loss": 0.4228,
+      "step": 42490
+    },
+    {
+      "epoch": 127.63,
+      "grad_norm": 14.823250770568848,
+      "learning_rate": 5.7457457457457466e-06,
+      "loss": 0.4349,
+      "step": 42500
+    },
+    {
+      "epoch": 127.66,
+      "grad_norm": 15.806150436401367,
+      "learning_rate": 5.744744744744746e-06,
+      "loss": 0.4243,
+      "step": 42510
+    },
+    {
+      "epoch": 127.69,
+      "grad_norm": 14.10430908203125,
+      "learning_rate": 5.743743743743744e-06,
+      "loss": 0.3898,
+      "step": 42520
+    },
+    {
+      "epoch": 127.72,
+      "grad_norm": 12.294818878173828,
+      "learning_rate": 5.742742742742743e-06,
+      "loss": 0.4963,
+      "step": 42530
+    },
+    {
+      "epoch": 127.75,
+      "grad_norm": 14.233175277709961,
+      "learning_rate": 5.741741741741742e-06,
+      "loss": 0.4445,
+      "step": 42540
+    },
+    {
+      "epoch": 127.78,
+      "grad_norm": 15.69285774230957,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 0.4711,
+      "step": 42550
+    },
+    {
+      "epoch": 127.81,
+      "grad_norm": 20.488258361816406,
+      "learning_rate": 5.739739739739741e-06,
+      "loss": 0.4276,
+      "step": 42560
+    },
+    {
+      "epoch": 127.84,
+      "grad_norm": 15.74963092803955,
+      "learning_rate": 5.738738738738739e-06,
+      "loss": 0.4557,
+      "step": 42570
+    },
+    {
+      "epoch": 127.87,
+      "grad_norm": 15.388700485229492,
+      "learning_rate": 5.737737737737739e-06,
+      "loss": 0.3936,
+      "step": 42580
+    },
+    {
+      "epoch": 127.9,
+      "grad_norm": 12.88378620147705,
+      "learning_rate": 5.736736736736737e-06,
+      "loss": 0.4036,
+      "step": 42590
+    },
+    {
+      "epoch": 127.93,
+      "grad_norm": 17.932510375976562,
+      "learning_rate": 5.735735735735736e-06,
+      "loss": 0.4637,
+      "step": 42600
+    },
+    {
+      "epoch": 127.96,
+      "grad_norm": 18.77720832824707,
+      "learning_rate": 5.7347347347347345e-06,
+      "loss": 0.456,
+      "step": 42610
+    },
+    {
+      "epoch": 127.99,
+      "grad_norm": 19.17095375061035,
+      "learning_rate": 5.733733733733734e-06,
+      "loss": 0.4633,
+      "step": 42620
+    },
+    {
+      "epoch": 128.0,
+      "eval_accuracy": 0.8637,
+      "eval_loss": 0.5130276083946228,
+      "eval_runtime": 13.087,
+      "eval_samples_per_second": 764.114,
+      "eval_steps_per_second": 3.056,
+      "step": 42624
+    },
+    {
+      "epoch": 128.02,
+      "grad_norm": 10.161124229431152,
+      "learning_rate": 5.732732732732733e-06,
+      "loss": 0.3205,
+      "step": 42630
+    },
+    {
+      "epoch": 128.05,
+      "grad_norm": 18.420995712280273,
+      "learning_rate": 5.731731731731732e-06,
+      "loss": 0.3965,
+      "step": 42640
+    },
+    {
+      "epoch": 128.08,
+      "grad_norm": 12.180721282958984,
+      "learning_rate": 5.7307307307307314e-06,
+      "loss": 0.4347,
+      "step": 42650
+    },
+    {
+      "epoch": 128.11,
+      "grad_norm": 19.56247329711914,
+      "learning_rate": 5.729729729729731e-06,
+      "loss": 0.4243,
+      "step": 42660
+    },
+    {
+      "epoch": 128.14,
+      "grad_norm": 13.910493850708008,
+      "learning_rate": 5.7287287287287295e-06,
+      "loss": 0.425,
+      "step": 42670
+    },
+    {
+      "epoch": 128.17,
+      "grad_norm": 15.725464820861816,
+      "learning_rate": 5.727727727727728e-06,
+      "loss": 0.4214,
+      "step": 42680
+    },
+    {
+      "epoch": 128.2,
+      "grad_norm": 14.830903053283691,
+      "learning_rate": 5.726726726726727e-06,
+      "loss": 0.4372,
+      "step": 42690
+    },
+    {
+      "epoch": 128.23,
+      "grad_norm": 19.751434326171875,
+      "learning_rate": 5.725725725725726e-06,
+      "loss": 0.3867,
+      "step": 42700
+    },
+    {
+      "epoch": 128.26,
+      "grad_norm": 13.786177635192871,
+      "learning_rate": 5.724724724724725e-06,
+      "loss": 0.4455,
+      "step": 42710
+    },
+    {
+      "epoch": 128.29,
+      "grad_norm": 16.02610969543457,
+      "learning_rate": 5.723723723723724e-06,
+      "loss": 0.4812,
+      "step": 42720
+    },
+    {
+      "epoch": 128.32,
+      "grad_norm": 21.80393409729004,
+      "learning_rate": 5.7227227227227236e-06,
+      "loss": 0.4335,
+      "step": 42730
+    },
+    {
+      "epoch": 128.35,
+      "grad_norm": 19.719316482543945,
+      "learning_rate": 5.721721721721722e-06,
+      "loss": 0.4536,
+      "step": 42740
+    },
+    {
+      "epoch": 128.38,
+      "grad_norm": 14.43315601348877,
+      "learning_rate": 5.720720720720722e-06,
+      "loss": 0.4384,
+      "step": 42750
+    },
+    {
+      "epoch": 128.41,
+      "grad_norm": 14.393265724182129,
+      "learning_rate": 5.719719719719721e-06,
+      "loss": 0.4057,
+      "step": 42760
+    },
+    {
+      "epoch": 128.44,
+      "grad_norm": 21.126739501953125,
+      "learning_rate": 5.718718718718719e-06,
+      "loss": 0.4262,
+      "step": 42770
+    },
+    {
+      "epoch": 128.47,
+      "grad_norm": 13.757997512817383,
+      "learning_rate": 5.717717717717718e-06,
+      "loss": 0.4382,
+      "step": 42780
+    },
+    {
+      "epoch": 128.5,
+      "grad_norm": 18.875782012939453,
+      "learning_rate": 5.716716716716717e-06,
+      "loss": 0.4325,
+      "step": 42790
+    },
+    {
+      "epoch": 128.53,
+      "grad_norm": 21.33592414855957,
+      "learning_rate": 5.715715715715716e-06,
+      "loss": 0.4807,
+      "step": 42800
+    },
+    {
+      "epoch": 128.56,
+      "grad_norm": 13.097156524658203,
+      "learning_rate": 5.714714714714715e-06,
+      "loss": 0.4296,
+      "step": 42810
+    },
+    {
+      "epoch": 128.59,
+      "grad_norm": 16.572519302368164,
+      "learning_rate": 5.713713713713714e-06,
+      "loss": 0.386,
+      "step": 42820
+    },
+    {
+      "epoch": 128.62,
+      "grad_norm": 12.109639167785645,
+      "learning_rate": 5.712712712712714e-06,
+      "loss": 0.3717,
+      "step": 42830
+    },
+    {
+      "epoch": 128.65,
+      "grad_norm": 26.960535049438477,
+      "learning_rate": 5.711711711711712e-06,
+      "loss": 0.4494,
+      "step": 42840
+    },
+    {
+      "epoch": 128.68,
+      "grad_norm": 20.343475341796875,
+      "learning_rate": 5.710710710710711e-06,
+      "loss": 0.4412,
+      "step": 42850
+    },
+    {
+      "epoch": 128.71,
+      "grad_norm": 18.60637664794922,
+      "learning_rate": 5.7097097097097095e-06,
+      "loss": 0.414,
+      "step": 42860
+    },
+    {
+      "epoch": 128.74,
+      "grad_norm": 14.138333320617676,
+      "learning_rate": 5.708708708708709e-06,
+      "loss": 0.3985,
+      "step": 42870
+    },
+    {
+      "epoch": 128.77,
+      "grad_norm": 15.098651885986328,
+      "learning_rate": 5.707707707707708e-06,
+      "loss": 0.3849,
+      "step": 42880
+    },
+    {
+      "epoch": 128.8,
+      "grad_norm": 19.921737670898438,
+      "learning_rate": 5.706706706706707e-06,
+      "loss": 0.391,
+      "step": 42890
+    },
+    {
+      "epoch": 128.83,
+      "grad_norm": 20.910551071166992,
+      "learning_rate": 5.7057057057057065e-06,
+      "loss": 0.4351,
+      "step": 42900
+    },
+    {
+      "epoch": 128.86,
+      "grad_norm": 22.095125198364258,
+      "learning_rate": 5.704704704704706e-06,
+      "loss": 0.4566,
+      "step": 42910
+    },
+    {
+      "epoch": 128.89,
+      "grad_norm": 10.166604042053223,
+      "learning_rate": 5.7037037037037045e-06,
+      "loss": 0.3953,
+      "step": 42920
+    },
+    {
+      "epoch": 128.92,
+      "grad_norm": 12.6982421875,
+      "learning_rate": 5.702702702702702e-06,
+      "loss": 0.4119,
+      "step": 42930
+    },
+    {
+      "epoch": 128.95,
+      "grad_norm": 15.591934204101562,
+      "learning_rate": 5.701701701701702e-06,
+      "loss": 0.444,
+      "step": 42940
+    },
+    {
+      "epoch": 128.98,
+      "grad_norm": 13.516217231750488,
+      "learning_rate": 5.700700700700701e-06,
+      "loss": 0.4259,
+      "step": 42950
+    },
+    {
+      "epoch": 129.0,
+      "eval_accuracy": 0.867,
+      "eval_loss": 0.5052863955497742,
+      "eval_runtime": 12.6818,
+      "eval_samples_per_second": 788.534,
+      "eval_steps_per_second": 3.154,
+      "step": 42957
+    },
+    {
+      "epoch": 129.01,
+      "grad_norm": 18.828712463378906,
+      "learning_rate": 5.6996996996997e-06,
+      "loss": 0.3751,
+      "step": 42960
+    },
+    {
+      "epoch": 129.04,
+      "grad_norm": 18.18568229675293,
+      "learning_rate": 5.698698698698699e-06,
+      "loss": 0.4188,
+      "step": 42970
+    },
+    {
+      "epoch": 129.07,
+      "grad_norm": 19.280569076538086,
+      "learning_rate": 5.697697697697699e-06,
+      "loss": 0.3914,
+      "step": 42980
+    },
+    {
+      "epoch": 129.1,
+      "grad_norm": 13.537618637084961,
+      "learning_rate": 5.696696696696697e-06,
+      "loss": 0.452,
+      "step": 42990
+    },
+    {
+      "epoch": 129.13,
+      "grad_norm": 10.954686164855957,
+      "learning_rate": 5.695695695695697e-06,
+      "loss": 0.4481,
+      "step": 43000
+    },
+    {
+      "epoch": 129.16,
+      "grad_norm": 12.040289878845215,
+      "learning_rate": 5.694694694694696e-06,
+      "loss": 0.3966,
+      "step": 43010
+    },
+    {
+      "epoch": 129.19,
+      "grad_norm": 11.69954776763916,
+      "learning_rate": 5.693693693693694e-06,
+      "loss": 0.4186,
+      "step": 43020
+    },
+    {
+      "epoch": 129.22,
+      "grad_norm": 17.634490966796875,
+      "learning_rate": 5.692692692692693e-06,
+      "loss": 0.4742,
+      "step": 43030
+    },
+    {
+      "epoch": 129.25,
+      "grad_norm": 18.072519302368164,
+      "learning_rate": 5.691691691691692e-06,
+      "loss": 0.3841,
+      "step": 43040
+    },
+    {
+      "epoch": 129.28,
+      "grad_norm": 15.875168800354004,
+      "learning_rate": 5.690690690690691e-06,
+      "loss": 0.4084,
+      "step": 43050
+    },
+    {
+      "epoch": 129.31,
+      "grad_norm": 13.685356140136719,
+      "learning_rate": 5.68968968968969e-06,
+      "loss": 0.4031,
+      "step": 43060
+    },
+    {
+      "epoch": 129.34,
+      "grad_norm": 16.941612243652344,
+      "learning_rate": 5.688688688688689e-06,
+      "loss": 0.4054,
+      "step": 43070
+    },
+    {
+      "epoch": 129.37,
+      "grad_norm": 7.8879594802856445,
+      "learning_rate": 5.687687687687689e-06,
+      "loss": 0.4157,
+      "step": 43080
+    },
+    {
+      "epoch": 129.4,
+      "grad_norm": 12.56670093536377,
+      "learning_rate": 5.686686686686687e-06,
+      "loss": 0.4064,
+      "step": 43090
+    },
+    {
+      "epoch": 129.43,
+      "grad_norm": 23.398212432861328,
+      "learning_rate": 5.685685685685686e-06,
+      "loss": 0.4523,
+      "step": 43100
+    },
+    {
+      "epoch": 129.46,
+      "grad_norm": 15.787158012390137,
+      "learning_rate": 5.6846846846846846e-06,
+      "loss": 0.4224,
+      "step": 43110
+    },
+    {
+      "epoch": 129.49,
+      "grad_norm": 9.929488182067871,
+      "learning_rate": 5.683683683683684e-06,
+      "loss": 0.4366,
+      "step": 43120
+    },
+    {
+      "epoch": 129.52,
+      "grad_norm": 23.603294372558594,
+      "learning_rate": 5.6826826826826835e-06,
+      "loss": 0.3901,
+      "step": 43130
+    },
+    {
+      "epoch": 129.55,
+      "grad_norm": 13.896695137023926,
+      "learning_rate": 5.681681681681682e-06,
+      "loss": 0.4424,
+      "step": 43140
+    },
+    {
+      "epoch": 129.58,
+      "grad_norm": 18.64598846435547,
+      "learning_rate": 5.6806806806806815e-06,
+      "loss": 0.4556,
+      "step": 43150
+    },
+    {
+      "epoch": 129.61,
+      "grad_norm": 14.568280220031738,
+      "learning_rate": 5.67967967967968e-06,
+      "loss": 0.405,
+      "step": 43160
+    },
+    {
+      "epoch": 129.64,
+      "grad_norm": 12.079632759094238,
+      "learning_rate": 5.6786786786786795e-06,
+      "loss": 0.4113,
+      "step": 43170
+    },
+    {
+      "epoch": 129.67,
+      "grad_norm": 13.248680114746094,
+      "learning_rate": 5.677677677677677e-06,
+      "loss": 0.471,
+      "step": 43180
+    },
+    {
+      "epoch": 129.7,
+      "grad_norm": 14.918341636657715,
+      "learning_rate": 5.676676676676677e-06,
+      "loss": 0.4236,
+      "step": 43190
+    },
+    {
+      "epoch": 129.73,
+      "grad_norm": 25.852066040039062,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.4411,
+      "step": 43200
+    },
+    {
+      "epoch": 129.76,
+      "grad_norm": 13.492530822753906,
+      "learning_rate": 5.674674674674675e-06,
+      "loss": 0.4295,
+      "step": 43210
+    },
+    {
+      "epoch": 129.79,
+      "grad_norm": 17.980485916137695,
+      "learning_rate": 5.673673673673674e-06,
+      "loss": 0.4466,
+      "step": 43220
+    },
+    {
+      "epoch": 129.82,
+      "grad_norm": 16.108491897583008,
+      "learning_rate": 5.672672672672674e-06,
+      "loss": 0.4121,
+      "step": 43230
+    },
+    {
+      "epoch": 129.85,
+      "grad_norm": 21.596431732177734,
+      "learning_rate": 5.671671671671672e-06,
+      "loss": 0.4454,
+      "step": 43240
+    },
+    {
+      "epoch": 129.88,
+      "grad_norm": 19.216691970825195,
+      "learning_rate": 5.670670670670672e-06,
+      "loss": 0.4565,
+      "step": 43250
+    },
+    {
+      "epoch": 129.91,
+      "grad_norm": 33.23293685913086,
+      "learning_rate": 5.6696696696696694e-06,
+      "loss": 0.4002,
+      "step": 43260
+    },
+    {
+      "epoch": 129.94,
+      "grad_norm": 11.879194259643555,
+      "learning_rate": 5.668668668668669e-06,
+      "loss": 0.4237,
+      "step": 43270
+    },
+    {
+      "epoch": 129.97,
+      "grad_norm": 18.151782989501953,
+      "learning_rate": 5.6676676676676675e-06,
+      "loss": 0.4216,
+      "step": 43280
+    },
+    {
+      "epoch": 130.0,
+      "grad_norm": 88.90621185302734,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 0.4668,
+      "step": 43290
+    },
+    {
+      "epoch": 130.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.5130959749221802,
+      "eval_runtime": 12.4923,
+      "eval_samples_per_second": 800.492,
+      "eval_steps_per_second": 3.202,
+      "step": 43290
+    },
+    {
+      "epoch": 130.03,
+      "grad_norm": 15.043937683105469,
+      "learning_rate": 5.665665665665666e-06,
+      "loss": 0.4371,
+      "step": 43300
+    },
+    {
+      "epoch": 130.06,
+      "grad_norm": 13.74203872680664,
+      "learning_rate": 5.664664664664665e-06,
+      "loss": 0.4327,
+      "step": 43310
+    },
+    {
+      "epoch": 130.09,
+      "grad_norm": 26.613210678100586,
+      "learning_rate": 5.663663663663664e-06,
+      "loss": 0.4081,
+      "step": 43320
+    },
+    {
+      "epoch": 130.12,
+      "grad_norm": 13.203279495239258,
+      "learning_rate": 5.662662662662664e-06,
+      "loss": 0.3979,
+      "step": 43330
+    },
+    {
+      "epoch": 130.15,
+      "grad_norm": 14.656771659851074,
+      "learning_rate": 5.661661661661662e-06,
+      "loss": 0.407,
+      "step": 43340
+    },
+    {
+      "epoch": 130.18,
+      "grad_norm": 13.669487953186035,
+      "learning_rate": 5.660660660660661e-06,
+      "loss": 0.3573,
+      "step": 43350
+    },
+    {
+      "epoch": 130.21,
+      "grad_norm": 13.50046443939209,
+      "learning_rate": 5.65965965965966e-06,
+      "loss": 0.3948,
+      "step": 43360
+    },
+    {
+      "epoch": 130.24,
+      "grad_norm": 15.160467147827148,
+      "learning_rate": 5.658658658658659e-06,
+      "loss": 0.3808,
+      "step": 43370
+    },
+    {
+      "epoch": 130.27,
+      "grad_norm": 16.645280838012695,
+      "learning_rate": 5.657657657657658e-06,
+      "loss": 0.3958,
+      "step": 43380
+    },
+    {
+      "epoch": 130.3,
+      "grad_norm": 16.091915130615234,
+      "learning_rate": 5.656656656656657e-06,
+      "loss": 0.4544,
+      "step": 43390
+    },
+    {
+      "epoch": 130.33,
+      "grad_norm": 14.931068420410156,
+      "learning_rate": 5.6556556556556565e-06,
+      "loss": 0.4828,
+      "step": 43400
+    },
+    {
+      "epoch": 130.36,
+      "grad_norm": 17.546337127685547,
+      "learning_rate": 5.654654654654655e-06,
+      "loss": 0.4553,
+      "step": 43410
+    },
+    {
+      "epoch": 130.39,
+      "grad_norm": 16.671566009521484,
+      "learning_rate": 5.6536536536536546e-06,
+      "loss": 0.4026,
+      "step": 43420
+    },
+    {
+      "epoch": 130.42,
+      "grad_norm": 22.365013122558594,
+      "learning_rate": 5.652652652652652e-06,
+      "loss": 0.4223,
+      "step": 43430
+    },
+    {
+      "epoch": 130.45,
+      "grad_norm": 15.041510581970215,
+      "learning_rate": 5.651651651651652e-06,
+      "loss": 0.3854,
+      "step": 43440
+    },
+    {
+      "epoch": 130.48,
+      "grad_norm": 15.137139320373535,
+      "learning_rate": 5.650650650650651e-06,
+      "loss": 0.4257,
+      "step": 43450
+    },
+    {
+      "epoch": 130.51,
+      "grad_norm": 16.203144073486328,
+      "learning_rate": 5.64964964964965e-06,
+      "loss": 0.4719,
+      "step": 43460
+    },
+    {
+      "epoch": 130.54,
+      "grad_norm": 15.88243293762207,
+      "learning_rate": 5.648648648648649e-06,
+      "loss": 0.3839,
+      "step": 43470
+    },
+    {
+      "epoch": 130.57,
+      "grad_norm": 16.286008834838867,
+      "learning_rate": 5.647647647647649e-06,
+      "loss": 0.4174,
+      "step": 43480
+    },
+    {
+      "epoch": 130.6,
+      "grad_norm": 12.633126258850098,
+      "learning_rate": 5.646646646646647e-06,
+      "loss": 0.4509,
+      "step": 43490
+    },
+    {
+      "epoch": 130.63,
+      "grad_norm": 14.958235740661621,
+      "learning_rate": 5.645645645645647e-06,
+      "loss": 0.4072,
+      "step": 43500
+    },
+    {
+      "epoch": 130.66,
+      "grad_norm": 15.095487594604492,
+      "learning_rate": 5.6446446446446445e-06,
+      "loss": 0.3578,
+      "step": 43510
+    },
+    {
+      "epoch": 130.69,
+      "grad_norm": 13.433402061462402,
+      "learning_rate": 5.643643643643644e-06,
+      "loss": 0.5075,
+      "step": 43520
+    },
+    {
+      "epoch": 130.72,
+      "grad_norm": 22.364097595214844,
+      "learning_rate": 5.6426426426426425e-06,
+      "loss": 0.4524,
+      "step": 43530
+    },
+    {
+      "epoch": 130.75,
+      "grad_norm": 28.61610221862793,
+      "learning_rate": 5.641641641641642e-06,
+      "loss": 0.4206,
+      "step": 43540
+    },
+    {
+      "epoch": 130.78,
+      "grad_norm": 14.025111198425293,
+      "learning_rate": 5.640640640640641e-06,
+      "loss": 0.3969,
+      "step": 43550
+    },
+    {
+      "epoch": 130.81,
+      "grad_norm": 16.909046173095703,
+      "learning_rate": 5.63963963963964e-06,
+      "loss": 0.418,
+      "step": 43560
+    },
+    {
+      "epoch": 130.84,
+      "grad_norm": 11.624141693115234,
+      "learning_rate": 5.638638638638639e-06,
+      "loss": 0.4128,
+      "step": 43570
+    },
+    {
+      "epoch": 130.87,
+      "grad_norm": 16.338165283203125,
+      "learning_rate": 5.637637637637639e-06,
+      "loss": 0.4216,
+      "step": 43580
+    },
+    {
+      "epoch": 130.9,
+      "grad_norm": 17.255868911743164,
+      "learning_rate": 5.6366366366366375e-06,
+      "loss": 0.4582,
+      "step": 43590
+    },
+    {
+      "epoch": 130.93,
+      "grad_norm": 16.18084716796875,
+      "learning_rate": 5.635635635635636e-06,
+      "loss": 0.4164,
+      "step": 43600
+    },
+    {
+      "epoch": 130.96,
+      "grad_norm": 12.81671142578125,
+      "learning_rate": 5.634634634634635e-06,
+      "loss": 0.4412,
+      "step": 43610
+    },
+    {
+      "epoch": 130.99,
+      "grad_norm": 12.275124549865723,
+      "learning_rate": 5.633633633633634e-06,
+      "loss": 0.4916,
+      "step": 43620
+    },
+    {
+      "epoch": 131.0,
+      "eval_accuracy": 0.8656,
+      "eval_loss": 0.5054722428321838,
+      "eval_runtime": 12.8379,
+      "eval_samples_per_second": 778.941,
+      "eval_steps_per_second": 3.116,
+      "step": 43623
+    },
+    {
+      "epoch": 131.02,
+      "grad_norm": 10.756814956665039,
+      "learning_rate": 5.632632632632633e-06,
+      "loss": 0.3898,
+      "step": 43630
+    },
+    {
+      "epoch": 131.05,
+      "grad_norm": 23.804399490356445,
+      "learning_rate": 5.631631631631632e-06,
+      "loss": 0.4845,
+      "step": 43640
+    },
+    {
+      "epoch": 131.08,
+      "grad_norm": 11.752922058105469,
+      "learning_rate": 5.6306306306306316e-06,
+      "loss": 0.4008,
+      "step": 43650
+    },
+    {
+      "epoch": 131.11,
+      "grad_norm": 12.261713981628418,
+      "learning_rate": 5.62962962962963e-06,
+      "loss": 0.4251,
+      "step": 43660
+    },
+    {
+      "epoch": 131.14,
+      "grad_norm": 21.01729393005371,
+      "learning_rate": 5.62862862862863e-06,
+      "loss": 0.4388,
+      "step": 43670
+    },
+    {
+      "epoch": 131.17,
+      "grad_norm": 23.794994354248047,
+      "learning_rate": 5.627627627627627e-06,
+      "loss": 0.3746,
+      "step": 43680
+    },
+    {
+      "epoch": 131.2,
+      "grad_norm": 11.116755485534668,
+      "learning_rate": 5.626626626626627e-06,
+      "loss": 0.399,
+      "step": 43690
+    },
+    {
+      "epoch": 131.23,
+      "grad_norm": 12.440905570983887,
+      "learning_rate": 5.625625625625626e-06,
+      "loss": 0.417,
+      "step": 43700
+    },
+    {
+      "epoch": 131.26,
+      "grad_norm": 23.417388916015625,
+      "learning_rate": 5.624624624624625e-06,
+      "loss": 0.4943,
+      "step": 43710
+    },
+    {
+      "epoch": 131.29,
+      "grad_norm": 13.447277069091797,
+      "learning_rate": 5.623623623623624e-06,
+      "loss": 0.4164,
+      "step": 43720
+    },
+    {
+      "epoch": 131.32,
+      "grad_norm": 16.68264389038086,
+      "learning_rate": 5.622622622622623e-06,
+      "loss": 0.4388,
+      "step": 43730
+    },
+    {
+      "epoch": 131.35,
+      "grad_norm": 12.662317276000977,
+      "learning_rate": 5.621621621621622e-06,
+      "loss": 0.3709,
+      "step": 43740
+    },
+    {
+      "epoch": 131.38,
+      "grad_norm": 15.70790958404541,
+      "learning_rate": 5.620620620620622e-06,
+      "loss": 0.4495,
+      "step": 43750
+    },
+    {
+      "epoch": 131.41,
+      "grad_norm": 11.975784301757812,
+      "learning_rate": 5.6196196196196195e-06,
+      "loss": 0.3897,
+      "step": 43760
+    },
+    {
+      "epoch": 131.44,
+      "grad_norm": 9.7977294921875,
+      "learning_rate": 5.618618618618619e-06,
+      "loss": 0.3844,
+      "step": 43770
+    },
+    {
+      "epoch": 131.47,
+      "grad_norm": 9.364118576049805,
+      "learning_rate": 5.6176176176176175e-06,
+      "loss": 0.4287,
+      "step": 43780
+    },
+    {
+      "epoch": 131.5,
+      "grad_norm": 14.867539405822754,
+      "learning_rate": 5.616616616616617e-06,
+      "loss": 0.4478,
+      "step": 43790
+    },
+    {
+      "epoch": 131.53,
+      "grad_norm": 10.32617473602295,
+      "learning_rate": 5.615615615615616e-06,
+      "loss": 0.3482,
+      "step": 43800
+    },
+    {
+      "epoch": 131.56,
+      "grad_norm": 19.557029724121094,
+      "learning_rate": 5.614614614614615e-06,
+      "loss": 0.4318,
+      "step": 43810
+    },
+    {
+      "epoch": 131.59,
+      "grad_norm": 14.144600868225098,
+      "learning_rate": 5.6136136136136145e-06,
+      "loss": 0.3982,
+      "step": 43820
+    },
+    {
+      "epoch": 131.62,
+      "grad_norm": 12.913371086120605,
+      "learning_rate": 5.612612612612614e-06,
+      "loss": 0.4024,
+      "step": 43830
+    },
+    {
+      "epoch": 131.65,
+      "grad_norm": 15.287763595581055,
+      "learning_rate": 5.6116116116116125e-06,
+      "loss": 0.4683,
+      "step": 43840
+    },
+    {
+      "epoch": 131.68,
+      "grad_norm": 16.151264190673828,
+      "learning_rate": 5.61061061061061e-06,
+      "loss": 0.4314,
+      "step": 43850
+    },
+    {
+      "epoch": 131.71,
+      "grad_norm": 15.09422492980957,
+      "learning_rate": 5.60960960960961e-06,
+      "loss": 0.4363,
+      "step": 43860
+    },
+    {
+      "epoch": 131.74,
+      "grad_norm": 22.04693603515625,
+      "learning_rate": 5.608608608608609e-06,
+      "loss": 0.4509,
+      "step": 43870
+    },
+    {
+      "epoch": 131.77,
+      "grad_norm": 16.221702575683594,
+      "learning_rate": 5.607607607607608e-06,
+      "loss": 0.4229,
+      "step": 43880
+    },
+    {
+      "epoch": 131.8,
+      "grad_norm": 9.259857177734375,
+      "learning_rate": 5.606606606606607e-06,
+      "loss": 0.4074,
+      "step": 43890
+    },
+    {
+      "epoch": 131.83,
+      "grad_norm": 12.36758804321289,
+      "learning_rate": 5.605605605605607e-06,
+      "loss": 0.402,
+      "step": 43900
+    },
+    {
+      "epoch": 131.86,
+      "grad_norm": 14.460962295532227,
+      "learning_rate": 5.604604604604605e-06,
+      "loss": 0.4125,
+      "step": 43910
+    },
+    {
+      "epoch": 131.89,
+      "grad_norm": 17.671276092529297,
+      "learning_rate": 5.603603603603605e-06,
+      "loss": 0.423,
+      "step": 43920
+    },
+    {
+      "epoch": 131.92,
+      "grad_norm": 15.976151466369629,
+      "learning_rate": 5.602602602602602e-06,
+      "loss": 0.4061,
+      "step": 43930
+    },
+    {
+      "epoch": 131.95,
+      "grad_norm": 22.482738494873047,
+      "learning_rate": 5.601601601601602e-06,
+      "loss": 0.4712,
+      "step": 43940
+    },
+    {
+      "epoch": 131.98,
+      "grad_norm": 19.360692977905273,
+      "learning_rate": 5.600600600600601e-06,
+      "loss": 0.4068,
+      "step": 43950
+    },
+    {
+      "epoch": 132.0,
+      "eval_accuracy": 0.8669,
+      "eval_loss": 0.5117104053497314,
+      "eval_runtime": 12.8393,
+      "eval_samples_per_second": 778.862,
+      "eval_steps_per_second": 3.115,
+      "step": 43956
+    },
+    {
+      "epoch": 132.01,
+      "grad_norm": 13.443767547607422,
+      "learning_rate": 5.5995995995996e-06,
+      "loss": 0.4507,
+      "step": 43960
+    },
+    {
+      "epoch": 132.04,
+      "grad_norm": 16.81865692138672,
+      "learning_rate": 5.598598598598599e-06,
+      "loss": 0.424,
+      "step": 43970
+    },
+    {
+      "epoch": 132.07,
+      "grad_norm": 12.607978820800781,
+      "learning_rate": 5.597597597597598e-06,
+      "loss": 0.3672,
+      "step": 43980
+    },
+    {
+      "epoch": 132.1,
+      "grad_norm": 12.531508445739746,
+      "learning_rate": 5.596596596596597e-06,
+      "loss": 0.4033,
+      "step": 43990
+    },
+    {
+      "epoch": 132.13,
+      "grad_norm": 25.236061096191406,
+      "learning_rate": 5.595595595595597e-06,
+      "loss": 0.438,
+      "step": 44000
+    },
+    {
+      "epoch": 132.16,
+      "grad_norm": 17.097007751464844,
+      "learning_rate": 5.5945945945945945e-06,
+      "loss": 0.3486,
+      "step": 44010
+    },
+    {
+      "epoch": 132.19,
+      "grad_norm": 14.821367263793945,
+      "learning_rate": 5.593593593593594e-06,
+      "loss": 0.398,
+      "step": 44020
+    },
+    {
+      "epoch": 132.22,
+      "grad_norm": 16.696918487548828,
+      "learning_rate": 5.5925925925925926e-06,
+      "loss": 0.4073,
+      "step": 44030
+    },
+    {
+      "epoch": 132.25,
+      "grad_norm": 17.518327713012695,
+      "learning_rate": 5.591591591591592e-06,
+      "loss": 0.4303,
+      "step": 44040
+    },
+    {
+      "epoch": 132.28,
+      "grad_norm": 14.253488540649414,
+      "learning_rate": 5.5905905905905915e-06,
+      "loss": 0.4279,
+      "step": 44050
+    },
+    {
+      "epoch": 132.31,
+      "grad_norm": 10.649962425231934,
+      "learning_rate": 5.58958958958959e-06,
+      "loss": 0.4175,
+      "step": 44060
+    },
+    {
+      "epoch": 132.34,
+      "grad_norm": 15.122641563415527,
+      "learning_rate": 5.5885885885885895e-06,
+      "loss": 0.4681,
+      "step": 44070
+    },
+    {
+      "epoch": 132.37,
+      "grad_norm": 26.50657844543457,
+      "learning_rate": 5.587587587587588e-06,
+      "loss": 0.3909,
+      "step": 44080
+    },
+    {
+      "epoch": 132.4,
+      "grad_norm": 20.504972457885742,
+      "learning_rate": 5.5865865865865875e-06,
+      "loss": 0.4406,
+      "step": 44090
+    },
+    {
+      "epoch": 132.43,
+      "grad_norm": 18.850160598754883,
+      "learning_rate": 5.585585585585585e-06,
+      "loss": 0.4422,
+      "step": 44100
+    },
+    {
+      "epoch": 132.46,
+      "grad_norm": 16.615657806396484,
+      "learning_rate": 5.584584584584585e-06,
+      "loss": 0.4643,
+      "step": 44110
+    },
+    {
+      "epoch": 132.49,
+      "grad_norm": 14.778960227966309,
+      "learning_rate": 5.583583583583584e-06,
+      "loss": 0.4253,
+      "step": 44120
+    },
+    {
+      "epoch": 132.52,
+      "grad_norm": 15.877119064331055,
+      "learning_rate": 5.582582582582583e-06,
+      "loss": 0.4217,
+      "step": 44130
+    },
+    {
+      "epoch": 132.55,
+      "grad_norm": 21.544187545776367,
+      "learning_rate": 5.581581581581582e-06,
+      "loss": 0.447,
+      "step": 44140
+    },
+    {
+      "epoch": 132.58,
+      "grad_norm": 12.980415344238281,
+      "learning_rate": 5.580580580580582e-06,
+      "loss": 0.4389,
+      "step": 44150
+    },
+    {
+      "epoch": 132.61,
+      "grad_norm": 17.50397300720215,
+      "learning_rate": 5.57957957957958e-06,
+      "loss": 0.3986,
+      "step": 44160
+    },
+    {
+      "epoch": 132.64,
+      "grad_norm": 13.513784408569336,
+      "learning_rate": 5.57857857857858e-06,
+      "loss": 0.4186,
+      "step": 44170
+    },
+    {
+      "epoch": 132.67,
+      "grad_norm": 15.283546447753906,
+      "learning_rate": 5.577577577577577e-06,
+      "loss": 0.4927,
+      "step": 44180
+    },
+    {
+      "epoch": 132.7,
+      "grad_norm": 16.641138076782227,
+      "learning_rate": 5.576576576576577e-06,
+      "loss": 0.4663,
+      "step": 44190
+    },
+    {
+      "epoch": 132.73,
+      "grad_norm": 18.672513961791992,
+      "learning_rate": 5.5755755755755755e-06,
+      "loss": 0.3808,
+      "step": 44200
+    },
+    {
+      "epoch": 132.76,
+      "grad_norm": 14.025522232055664,
+      "learning_rate": 5.574574574574575e-06,
+      "loss": 0.4615,
+      "step": 44210
+    },
+    {
+      "epoch": 132.79,
+      "grad_norm": 20.64201545715332,
+      "learning_rate": 5.573573573573574e-06,
+      "loss": 0.4436,
+      "step": 44220
+    },
+    {
+      "epoch": 132.82,
+      "grad_norm": 21.664323806762695,
+      "learning_rate": 5.572572572572573e-06,
+      "loss": 0.3882,
+      "step": 44230
+    },
+    {
+      "epoch": 132.85,
+      "grad_norm": 19.86939811706543,
+      "learning_rate": 5.571571571571572e-06,
+      "loss": 0.4238,
+      "step": 44240
+    },
+    {
+      "epoch": 132.88,
+      "grad_norm": 16.02164649963379,
+      "learning_rate": 5.570570570570572e-06,
+      "loss": 0.4255,
+      "step": 44250
+    },
+    {
+      "epoch": 132.91,
+      "grad_norm": 13.511465072631836,
+      "learning_rate": 5.5695695695695696e-06,
+      "loss": 0.42,
+      "step": 44260
+    },
+    {
+      "epoch": 132.94,
+      "grad_norm": 11.168133735656738,
+      "learning_rate": 5.568568568568569e-06,
+      "loss": 0.479,
+      "step": 44270
+    },
+    {
+      "epoch": 132.97,
+      "grad_norm": 14.723389625549316,
+      "learning_rate": 5.567567567567568e-06,
+      "loss": 0.4187,
+      "step": 44280
+    },
+    {
+      "epoch": 133.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.5150870084762573,
+      "eval_runtime": 12.6551,
+      "eval_samples_per_second": 790.195,
+      "eval_steps_per_second": 3.161,
+      "step": 44289
+    },
+    {
+      "epoch": 133.0,
+      "grad_norm": 25.311315536499023,
+      "learning_rate": 5.566566566566567e-06,
+      "loss": 0.3993,
+      "step": 44290
+    },
+    {
+      "epoch": 133.03,
+      "grad_norm": 13.01799201965332,
+      "learning_rate": 5.565565565565566e-06,
+      "loss": 0.4122,
+      "step": 44300
+    },
+    {
+      "epoch": 133.06,
+      "grad_norm": 14.802847862243652,
+      "learning_rate": 5.564564564564565e-06,
+      "loss": 0.4,
+      "step": 44310
+    },
+    {
+      "epoch": 133.09,
+      "grad_norm": 18.1865177154541,
+      "learning_rate": 5.5635635635635645e-06,
+      "loss": 0.4015,
+      "step": 44320
+    },
+    {
+      "epoch": 133.12,
+      "grad_norm": 12.796468734741211,
+      "learning_rate": 5.562562562562563e-06,
+      "loss": 0.4548,
+      "step": 44330
+    },
+    {
+      "epoch": 133.15,
+      "grad_norm": 12.758103370666504,
+      "learning_rate": 5.561561561561562e-06,
+      "loss": 0.4206,
+      "step": 44340
+    },
+    {
+      "epoch": 133.18,
+      "grad_norm": 10.398483276367188,
+      "learning_rate": 5.56056056056056e-06,
+      "loss": 0.377,
+      "step": 44350
+    },
+    {
+      "epoch": 133.21,
+      "grad_norm": 15.863533973693848,
+      "learning_rate": 5.55955955955956e-06,
+      "loss": 0.4056,
+      "step": 44360
+    },
+    {
+      "epoch": 133.24,
+      "grad_norm": 18.868867874145508,
+      "learning_rate": 5.558558558558559e-06,
+      "loss": 0.4043,
+      "step": 44370
+    },
+    {
+      "epoch": 133.27,
+      "grad_norm": 19.306272506713867,
+      "learning_rate": 5.557557557557558e-06,
+      "loss": 0.4357,
+      "step": 44380
+    },
+    {
+      "epoch": 133.3,
+      "grad_norm": 17.742401123046875,
+      "learning_rate": 5.556556556556557e-06,
+      "loss": 0.4262,
+      "step": 44390
+    },
+    {
+      "epoch": 133.33,
+      "grad_norm": 16.406770706176758,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.3879,
+      "step": 44400
+    },
+    {
+      "epoch": 133.36,
+      "grad_norm": 14.064252853393555,
+      "learning_rate": 5.554554554554555e-06,
+      "loss": 0.4229,
+      "step": 44410
+    },
+    {
+      "epoch": 133.39,
+      "grad_norm": 13.283167839050293,
+      "learning_rate": 5.553553553553555e-06,
+      "loss": 0.4257,
+      "step": 44420
+    },
+    {
+      "epoch": 133.42,
+      "grad_norm": 18.681957244873047,
+      "learning_rate": 5.5525525525525525e-06,
+      "loss": 0.4402,
+      "step": 44430
+    },
+    {
+      "epoch": 133.45,
+      "grad_norm": 19.141603469848633,
+      "learning_rate": 5.551551551551552e-06,
+      "loss": 0.4143,
+      "step": 44440
+    },
+    {
+      "epoch": 133.48,
+      "grad_norm": 17.14235496520996,
+      "learning_rate": 5.5505505505505505e-06,
+      "loss": 0.3329,
+      "step": 44450
+    },
+    {
+      "epoch": 133.51,
+      "grad_norm": 12.723823547363281,
+      "learning_rate": 5.54954954954955e-06,
+      "loss": 0.4342,
+      "step": 44460
+    },
+    {
+      "epoch": 133.54,
+      "grad_norm": 20.307344436645508,
+      "learning_rate": 5.548548548548549e-06,
+      "loss": 0.4433,
+      "step": 44470
+    },
+    {
+      "epoch": 133.57,
+      "grad_norm": 20.673120498657227,
+      "learning_rate": 5.547547547547548e-06,
+      "loss": 0.4192,
+      "step": 44480
+    },
+    {
+      "epoch": 133.6,
+      "grad_norm": 15.02798080444336,
+      "learning_rate": 5.546546546546547e-06,
+      "loss": 0.38,
+      "step": 44490
+    },
+    {
+      "epoch": 133.63,
+      "grad_norm": 16.095252990722656,
+      "learning_rate": 5.545545545545547e-06,
+      "loss": 0.3643,
+      "step": 44500
+    },
+    {
+      "epoch": 133.66,
+      "grad_norm": 18.823829650878906,
+      "learning_rate": 5.544544544544545e-06,
+      "loss": 0.4032,
+      "step": 44510
+    },
+    {
+      "epoch": 133.69,
+      "grad_norm": 19.04279136657715,
+      "learning_rate": 5.543543543543544e-06,
+      "loss": 0.3621,
+      "step": 44520
+    },
+    {
+      "epoch": 133.72,
+      "grad_norm": 17.993934631347656,
+      "learning_rate": 5.542542542542543e-06,
+      "loss": 0.4014,
+      "step": 44530
+    },
+    {
+      "epoch": 133.75,
+      "grad_norm": 18.556066513061523,
+      "learning_rate": 5.541541541541542e-06,
+      "loss": 0.3815,
+      "step": 44540
+    },
+    {
+      "epoch": 133.78,
+      "grad_norm": 7.523036479949951,
+      "learning_rate": 5.540540540540541e-06,
+      "loss": 0.3871,
+      "step": 44550
+    },
+    {
+      "epoch": 133.81,
+      "grad_norm": 17.561464309692383,
+      "learning_rate": 5.53953953953954e-06,
+      "loss": 0.5137,
+      "step": 44560
+    },
+    {
+      "epoch": 133.84,
+      "grad_norm": 11.84272289276123,
+      "learning_rate": 5.5385385385385396e-06,
+      "loss": 0.3856,
+      "step": 44570
+    },
+    {
+      "epoch": 133.87,
+      "grad_norm": 17.587617874145508,
+      "learning_rate": 5.537537537537538e-06,
+      "loss": 0.4416,
+      "step": 44580
+    },
+    {
+      "epoch": 133.9,
+      "grad_norm": 20.27515983581543,
+      "learning_rate": 5.536536536536537e-06,
+      "loss": 0.432,
+      "step": 44590
+    },
+    {
+      "epoch": 133.93,
+      "grad_norm": 10.16312313079834,
+      "learning_rate": 5.535535535535535e-06,
+      "loss": 0.4076,
+      "step": 44600
+    },
+    {
+      "epoch": 133.96,
+      "grad_norm": 15.362261772155762,
+      "learning_rate": 5.534534534534535e-06,
+      "loss": 0.4033,
+      "step": 44610
+    },
+    {
+      "epoch": 133.99,
+      "grad_norm": 16.463560104370117,
+      "learning_rate": 5.533533533533534e-06,
+      "loss": 0.4197,
+      "step": 44620
+    },
+    {
+      "epoch": 134.0,
+      "eval_accuracy": 0.8685,
+      "eval_loss": 0.506835401058197,
+      "eval_runtime": 12.6946,
+      "eval_samples_per_second": 787.739,
+      "eval_steps_per_second": 3.151,
+      "step": 44622
+    },
+    {
+      "epoch": 134.02,
+      "grad_norm": 12.834206581115723,
+      "learning_rate": 5.532532532532533e-06,
+      "loss": 0.5647,
+      "step": 44630
+    },
+    {
+      "epoch": 134.05,
+      "grad_norm": 13.70963191986084,
+      "learning_rate": 5.531531531531532e-06,
+      "loss": 0.4123,
+      "step": 44640
+    },
+    {
+      "epoch": 134.08,
+      "grad_norm": 14.262741088867188,
+      "learning_rate": 5.530530530530531e-06,
+      "loss": 0.4241,
+      "step": 44650
+    },
+    {
+      "epoch": 134.11,
+      "grad_norm": 13.543120384216309,
+      "learning_rate": 5.52952952952953e-06,
+      "loss": 0.4077,
+      "step": 44660
+    },
+    {
+      "epoch": 134.14,
+      "grad_norm": 20.906463623046875,
+      "learning_rate": 5.52852852852853e-06,
+      "loss": 0.4079,
+      "step": 44670
+    },
+    {
+      "epoch": 134.17,
+      "grad_norm": 15.345733642578125,
+      "learning_rate": 5.5275275275275275e-06,
+      "loss": 0.429,
+      "step": 44680
+    },
+    {
+      "epoch": 134.2,
+      "grad_norm": 16.24897003173828,
+      "learning_rate": 5.526526526526527e-06,
+      "loss": 0.3655,
+      "step": 44690
+    },
+    {
+      "epoch": 134.23,
+      "grad_norm": 20.236555099487305,
+      "learning_rate": 5.5255255255255255e-06,
+      "loss": 0.4075,
+      "step": 44700
+    },
+    {
+      "epoch": 134.26,
+      "grad_norm": 16.102468490600586,
+      "learning_rate": 5.524524524524525e-06,
+      "loss": 0.4517,
+      "step": 44710
+    },
+    {
+      "epoch": 134.29,
+      "grad_norm": 14.781170845031738,
+      "learning_rate": 5.523523523523524e-06,
+      "loss": 0.3793,
+      "step": 44720
+    },
+    {
+      "epoch": 134.32,
+      "grad_norm": 19.067134857177734,
+      "learning_rate": 5.522522522522523e-06,
+      "loss": 0.4435,
+      "step": 44730
+    },
+    {
+      "epoch": 134.35,
+      "grad_norm": 10.715248107910156,
+      "learning_rate": 5.5215215215215224e-06,
+      "loss": 0.4406,
+      "step": 44740
+    },
+    {
+      "epoch": 134.38,
+      "grad_norm": 25.181217193603516,
+      "learning_rate": 5.520520520520521e-06,
+      "loss": 0.4258,
+      "step": 44750
+    },
+    {
+      "epoch": 134.41,
+      "grad_norm": 11.163804054260254,
+      "learning_rate": 5.51951951951952e-06,
+      "loss": 0.347,
+      "step": 44760
+    },
+    {
+      "epoch": 134.44,
+      "grad_norm": 18.208724975585938,
+      "learning_rate": 5.518518518518518e-06,
+      "loss": 0.3901,
+      "step": 44770
+    },
+    {
+      "epoch": 134.47,
+      "grad_norm": 12.728507995605469,
+      "learning_rate": 5.517517517517518e-06,
+      "loss": 0.4032,
+      "step": 44780
+    },
+    {
+      "epoch": 134.5,
+      "grad_norm": 17.121353149414062,
+      "learning_rate": 5.516516516516517e-06,
+      "loss": 0.4423,
+      "step": 44790
+    },
+    {
+      "epoch": 134.53,
+      "grad_norm": 16.886699676513672,
+      "learning_rate": 5.515515515515516e-06,
+      "loss": 0.477,
+      "step": 44800
+    },
+    {
+      "epoch": 134.56,
+      "grad_norm": 13.860877990722656,
+      "learning_rate": 5.514514514514515e-06,
+      "loss": 0.4744,
+      "step": 44810
+    },
+    {
+      "epoch": 134.59,
+      "grad_norm": 14.356751441955566,
+      "learning_rate": 5.513513513513515e-06,
+      "loss": 0.3983,
+      "step": 44820
+    },
+    {
+      "epoch": 134.62,
+      "grad_norm": 14.646833419799805,
+      "learning_rate": 5.512512512512513e-06,
+      "loss": 0.3775,
+      "step": 44830
+    },
+    {
+      "epoch": 134.65,
+      "grad_norm": 23.273635864257812,
+      "learning_rate": 5.511511511511512e-06,
+      "loss": 0.456,
+      "step": 44840
+    },
+    {
+      "epoch": 134.68,
+      "grad_norm": 17.371379852294922,
+      "learning_rate": 5.51051051051051e-06,
+      "loss": 0.4423,
+      "step": 44850
+    },
+    {
+      "epoch": 134.71,
+      "grad_norm": 12.991501808166504,
+      "learning_rate": 5.50950950950951e-06,
+      "loss": 0.4093,
+      "step": 44860
+    },
+    {
+      "epoch": 134.74,
+      "grad_norm": 16.685895919799805,
+      "learning_rate": 5.508508508508508e-06,
+      "loss": 0.4148,
+      "step": 44870
+    },
+    {
+      "epoch": 134.77,
+      "grad_norm": 17.519512176513672,
+      "learning_rate": 5.507507507507508e-06,
+      "loss": 0.4055,
+      "step": 44880
+    },
+    {
+      "epoch": 134.8,
+      "grad_norm": 23.86377716064453,
+      "learning_rate": 5.506506506506507e-06,
+      "loss": 0.407,
+      "step": 44890
+    },
+    {
+      "epoch": 134.83,
+      "grad_norm": 17.091838836669922,
+      "learning_rate": 5.505505505505506e-06,
+      "loss": 0.4829,
+      "step": 44900
+    },
+    {
+      "epoch": 134.86,
+      "grad_norm": 15.923513412475586,
+      "learning_rate": 5.504504504504505e-06,
+      "loss": 0.4164,
+      "step": 44910
+    },
+    {
+      "epoch": 134.89,
+      "grad_norm": 12.826001167297363,
+      "learning_rate": 5.503503503503505e-06,
+      "loss": 0.3411,
+      "step": 44920
+    },
+    {
+      "epoch": 134.92,
+      "grad_norm": 15.110023498535156,
+      "learning_rate": 5.5025025025025025e-06,
+      "loss": 0.4124,
+      "step": 44930
+    },
+    {
+      "epoch": 134.95,
+      "grad_norm": 17.33800506591797,
+      "learning_rate": 5.501501501501502e-06,
+      "loss": 0.4376,
+      "step": 44940
+    },
+    {
+      "epoch": 134.98,
+      "grad_norm": 12.34097957611084,
+      "learning_rate": 5.5005005005005006e-06,
+      "loss": 0.3916,
+      "step": 44950
+    },
+    {
+      "epoch": 135.0,
+      "eval_accuracy": 0.8684,
+      "eval_loss": 0.5028141140937805,
+      "eval_runtime": 12.7265,
+      "eval_samples_per_second": 785.759,
+      "eval_steps_per_second": 3.143,
+      "step": 44955
+    },
+    {
+      "epoch": 135.02,
+      "grad_norm": 13.759685516357422,
+      "learning_rate": 5.4994994994995e-06,
+      "loss": 0.4184,
+      "step": 44960
+    },
+    {
+      "epoch": 135.05,
+      "grad_norm": 13.3011474609375,
+      "learning_rate": 5.4984984984984994e-06,
+      "loss": 0.4337,
+      "step": 44970
+    },
+    {
+      "epoch": 135.08,
+      "grad_norm": 19.68037223815918,
+      "learning_rate": 5.497497497497498e-06,
+      "loss": 0.4179,
+      "step": 44980
+    },
+    {
+      "epoch": 135.11,
+      "grad_norm": 21.39244270324707,
+      "learning_rate": 5.4964964964964975e-06,
+      "loss": 0.4508,
+      "step": 44990
+    },
+    {
+      "epoch": 135.14,
+      "grad_norm": 15.901956558227539,
+      "learning_rate": 5.495495495495496e-06,
+      "loss": 0.4152,
+      "step": 45000
+    },
+    {
+      "epoch": 135.17,
+      "grad_norm": 15.306147575378418,
+      "learning_rate": 5.494494494494495e-06,
+      "loss": 0.4129,
+      "step": 45010
+    },
+    {
+      "epoch": 135.2,
+      "grad_norm": 11.089850425720215,
+      "learning_rate": 5.493493493493493e-06,
+      "loss": 0.4246,
+      "step": 45020
+    },
+    {
+      "epoch": 135.23,
+      "grad_norm": 16.303211212158203,
+      "learning_rate": 5.492492492492493e-06,
+      "loss": 0.3883,
+      "step": 45030
+    },
+    {
+      "epoch": 135.26,
+      "grad_norm": 18.1466007232666,
+      "learning_rate": 5.491491491491492e-06,
+      "loss": 0.4571,
+      "step": 45040
+    },
+    {
+      "epoch": 135.29,
+      "grad_norm": 14.853670120239258,
+      "learning_rate": 5.490490490490491e-06,
+      "loss": 0.4266,
+      "step": 45050
+    },
+    {
+      "epoch": 135.32,
+      "grad_norm": 19.884000778198242,
+      "learning_rate": 5.48948948948949e-06,
+      "loss": 0.4073,
+      "step": 45060
+    },
+    {
+      "epoch": 135.35,
+      "grad_norm": 14.912625312805176,
+      "learning_rate": 5.48848848848849e-06,
+      "loss": 0.3801,
+      "step": 45070
+    },
+    {
+      "epoch": 135.38,
+      "grad_norm": 15.538232803344727,
+      "learning_rate": 5.487487487487488e-06,
+      "loss": 0.3563,
+      "step": 45080
+    },
+    {
+      "epoch": 135.41,
+      "grad_norm": 14.936857223510742,
+      "learning_rate": 5.486486486486487e-06,
+      "loss": 0.4377,
+      "step": 45090
+    },
+    {
+      "epoch": 135.44,
+      "grad_norm": 16.08387565612793,
+      "learning_rate": 5.485485485485485e-06,
+      "loss": 0.4232,
+      "step": 45100
+    },
+    {
+      "epoch": 135.47,
+      "grad_norm": 22.15740203857422,
+      "learning_rate": 5.484484484484485e-06,
+      "loss": 0.4536,
+      "step": 45110
+    },
+    {
+      "epoch": 135.5,
+      "grad_norm": 14.446589469909668,
+      "learning_rate": 5.4834834834834834e-06,
+      "loss": 0.3485,
+      "step": 45120
+    },
+    {
+      "epoch": 135.53,
+      "grad_norm": 14.768562316894531,
+      "learning_rate": 5.482482482482483e-06,
+      "loss": 0.4762,
+      "step": 45130
+    },
+    {
+      "epoch": 135.56,
+      "grad_norm": 21.051868438720703,
+      "learning_rate": 5.481481481481482e-06,
+      "loss": 0.4058,
+      "step": 45140
+    },
+    {
+      "epoch": 135.59,
+      "grad_norm": 9.157076835632324,
+      "learning_rate": 5.480480480480481e-06,
+      "loss": 0.3726,
+      "step": 45150
+    },
+    {
+      "epoch": 135.62,
+      "grad_norm": 13.095471382141113,
+      "learning_rate": 5.47947947947948e-06,
+      "loss": 0.369,
+      "step": 45160
+    },
+    {
+      "epoch": 135.65,
+      "grad_norm": 18.004135131835938,
+      "learning_rate": 5.47847847847848e-06,
+      "loss": 0.4214,
+      "step": 45170
+    },
+    {
+      "epoch": 135.68,
+      "grad_norm": 18.944307327270508,
+      "learning_rate": 5.4774774774774776e-06,
+      "loss": 0.3743,
+      "step": 45180
+    },
+    {
+      "epoch": 135.71,
+      "grad_norm": 11.962390899658203,
+      "learning_rate": 5.476476476476477e-06,
+      "loss": 0.4089,
+      "step": 45190
+    },
+    {
+      "epoch": 135.74,
+      "grad_norm": 13.281852722167969,
+      "learning_rate": 5.475475475475476e-06,
+      "loss": 0.3993,
+      "step": 45200
+    },
+    {
+      "epoch": 135.77,
+      "grad_norm": 16.407140731811523,
+      "learning_rate": 5.474474474474475e-06,
+      "loss": 0.4299,
+      "step": 45210
+    },
+    {
+      "epoch": 135.8,
+      "grad_norm": 14.726142883300781,
+      "learning_rate": 5.473473473473474e-06,
+      "loss": 0.4432,
+      "step": 45220
+    },
+    {
+      "epoch": 135.83,
+      "grad_norm": 26.71658706665039,
+      "learning_rate": 5.472472472472473e-06,
+      "loss": 0.3987,
+      "step": 45230
+    },
+    {
+      "epoch": 135.86,
+      "grad_norm": 15.686315536499023,
+      "learning_rate": 5.4714714714714725e-06,
+      "loss": 0.401,
+      "step": 45240
+    },
+    {
+      "epoch": 135.89,
+      "grad_norm": 13.584820747375488,
+      "learning_rate": 5.470470470470471e-06,
+      "loss": 0.4046,
+      "step": 45250
+    },
+    {
+      "epoch": 135.92,
+      "grad_norm": 15.129951477050781,
+      "learning_rate": 5.46946946946947e-06,
+      "loss": 0.4201,
+      "step": 45260
+    },
+    {
+      "epoch": 135.95,
+      "grad_norm": 14.49087142944336,
+      "learning_rate": 5.468468468468468e-06,
+      "loss": 0.3892,
+      "step": 45270
+    },
+    {
+      "epoch": 135.98,
+      "grad_norm": 18.390094757080078,
+      "learning_rate": 5.467467467467468e-06,
+      "loss": 0.4084,
+      "step": 45280
+    },
+    {
+      "epoch": 136.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.5096805691719055,
+      "eval_runtime": 13.2317,
+      "eval_samples_per_second": 755.758,
+      "eval_steps_per_second": 3.023,
+      "step": 45288
+    },
+    {
+      "epoch": 136.01,
+      "grad_norm": 13.510574340820312,
+      "learning_rate": 5.466466466466467e-06,
+      "loss": 0.3638,
+      "step": 45290
+    },
+    {
+      "epoch": 136.04,
+      "grad_norm": 22.62605857849121,
+      "learning_rate": 5.465465465465466e-06,
+      "loss": 0.4337,
+      "step": 45300
+    },
+    {
+      "epoch": 136.07,
+      "grad_norm": 13.755040168762207,
+      "learning_rate": 5.464464464464465e-06,
+      "loss": 0.4345,
+      "step": 45310
+    },
+    {
+      "epoch": 136.1,
+      "grad_norm": 16.296117782592773,
+      "learning_rate": 5.463463463463464e-06,
+      "loss": 0.4055,
+      "step": 45320
+    },
+    {
+      "epoch": 136.13,
+      "grad_norm": 9.672499656677246,
+      "learning_rate": 5.462462462462463e-06,
+      "loss": 0.446,
+      "step": 45330
+    },
+    {
+      "epoch": 136.16,
+      "grad_norm": 13.168482780456543,
+      "learning_rate": 5.461461461461461e-06,
+      "loss": 0.428,
+      "step": 45340
+    },
+    {
+      "epoch": 136.19,
+      "grad_norm": 15.795666694641113,
+      "learning_rate": 5.4604604604604604e-06,
+      "loss": 0.4274,
+      "step": 45350
+    },
+    {
+      "epoch": 136.22,
+      "grad_norm": 11.331927299499512,
+      "learning_rate": 5.45945945945946e-06,
+      "loss": 0.4401,
+      "step": 45360
+    },
+    {
+      "epoch": 136.25,
+      "grad_norm": 16.028980255126953,
+      "learning_rate": 5.4584584584584585e-06,
+      "loss": 0.4347,
+      "step": 45370
+    },
+    {
+      "epoch": 136.28,
+      "grad_norm": 15.187689781188965,
+      "learning_rate": 5.457457457457458e-06,
+      "loss": 0.4116,
+      "step": 45380
+    },
+    {
+      "epoch": 136.31,
+      "grad_norm": 22.916255950927734,
+      "learning_rate": 5.456456456456457e-06,
+      "loss": 0.3711,
+      "step": 45390
+    },
+    {
+      "epoch": 136.34,
+      "grad_norm": 18.421016693115234,
+      "learning_rate": 5.455455455455456e-06,
+      "loss": 0.448,
+      "step": 45400
+    },
+    {
+      "epoch": 136.37,
+      "grad_norm": 16.356000900268555,
+      "learning_rate": 5.454454454454455e-06,
+      "loss": 0.4275,
+      "step": 45410
+    },
+    {
+      "epoch": 136.4,
+      "grad_norm": 14.906598091125488,
+      "learning_rate": 5.453453453453455e-06,
+      "loss": 0.464,
+      "step": 45420
+    },
+    {
+      "epoch": 136.43,
+      "grad_norm": 18.145944595336914,
+      "learning_rate": 5.452452452452453e-06,
+      "loss": 0.4439,
+      "step": 45430
+    },
+    {
+      "epoch": 136.46,
+      "grad_norm": 13.597373008728027,
+      "learning_rate": 5.451451451451451e-06,
+      "loss": 0.4464,
+      "step": 45440
+    },
+    {
+      "epoch": 136.49,
+      "grad_norm": 18.007658004760742,
+      "learning_rate": 5.450450450450451e-06,
+      "loss": 0.4038,
+      "step": 45450
+    },
+    {
+      "epoch": 136.52,
+      "grad_norm": 16.391271591186523,
+      "learning_rate": 5.44944944944945e-06,
+      "loss": 0.4364,
+      "step": 45460
+    },
+    {
+      "epoch": 136.55,
+      "grad_norm": 17.620119094848633,
+      "learning_rate": 5.448448448448449e-06,
+      "loss": 0.3709,
+      "step": 45470
+    },
+    {
+      "epoch": 136.58,
+      "grad_norm": 13.822389602661133,
+      "learning_rate": 5.447447447447448e-06,
+      "loss": 0.3732,
+      "step": 45480
+    },
+    {
+      "epoch": 136.61,
+      "grad_norm": 20.174545288085938,
+      "learning_rate": 5.4464464464464475e-06,
+      "loss": 0.4764,
+      "step": 45490
+    },
+    {
+      "epoch": 136.64,
+      "grad_norm": 11.752460479736328,
+      "learning_rate": 5.445445445445446e-06,
+      "loss": 0.355,
+      "step": 45500
+    },
+    {
+      "epoch": 136.67,
+      "grad_norm": 14.59684944152832,
+      "learning_rate": 5.444444444444445e-06,
+      "loss": 0.3812,
+      "step": 45510
+    },
+    {
+      "epoch": 136.7,
+      "grad_norm": 17.308778762817383,
+      "learning_rate": 5.443443443443443e-06,
+      "loss": 0.3919,
+      "step": 45520
+    },
+    {
+      "epoch": 136.73,
+      "grad_norm": 18.466934204101562,
+      "learning_rate": 5.442442442442443e-06,
+      "loss": 0.3717,
+      "step": 45530
+    },
+    {
+      "epoch": 136.76,
+      "grad_norm": 18.733753204345703,
+      "learning_rate": 5.441441441441442e-06,
+      "loss": 0.4339,
+      "step": 45540
+    },
+    {
+      "epoch": 136.79,
+      "grad_norm": 17.155197143554688,
+      "learning_rate": 5.440440440440441e-06,
+      "loss": 0.4429,
+      "step": 45550
+    },
+    {
+      "epoch": 136.82,
+      "grad_norm": 17.050233840942383,
+      "learning_rate": 5.43943943943944e-06,
+      "loss": 0.4122,
+      "step": 45560
+    },
+    {
+      "epoch": 136.85,
+      "grad_norm": 12.591126441955566,
+      "learning_rate": 5.438438438438439e-06,
+      "loss": 0.4379,
+      "step": 45570
+    },
+    {
+      "epoch": 136.88,
+      "grad_norm": 18.957799911499023,
+      "learning_rate": 5.437437437437438e-06,
+      "loss": 0.4551,
+      "step": 45580
+    },
+    {
+      "epoch": 136.91,
+      "grad_norm": 14.412091255187988,
+      "learning_rate": 5.436436436436436e-06,
+      "loss": 0.4306,
+      "step": 45590
+    },
+    {
+      "epoch": 136.94,
+      "grad_norm": 14.689226150512695,
+      "learning_rate": 5.4354354354354355e-06,
+      "loss": 0.4177,
+      "step": 45600
+    },
+    {
+      "epoch": 136.97,
+      "grad_norm": 12.581292152404785,
+      "learning_rate": 5.434434434434435e-06,
+      "loss": 0.4014,
+      "step": 45610
+    },
+    {
+      "epoch": 137.0,
+      "grad_norm": 12.538034439086914,
+      "learning_rate": 5.4334334334334335e-06,
+      "loss": 0.405,
+      "step": 45620
+    },
+    {
+      "epoch": 137.0,
+      "eval_accuracy": 0.8664,
+      "eval_loss": 0.5061039328575134,
+      "eval_runtime": 12.7126,
+      "eval_samples_per_second": 786.624,
+      "eval_steps_per_second": 3.146,
+      "step": 45621
+    },
+    {
+      "epoch": 137.03,
+      "grad_norm": 19.187183380126953,
+      "learning_rate": 5.432432432432433e-06,
+      "loss": 0.3527,
+      "step": 45630
+    },
+    {
+      "epoch": 137.06,
+      "grad_norm": 14.550982475280762,
+      "learning_rate": 5.431431431431432e-06,
+      "loss": 0.4064,
+      "step": 45640
+    },
+    {
+      "epoch": 137.09,
+      "grad_norm": 17.989953994750977,
+      "learning_rate": 5.430430430430431e-06,
+      "loss": 0.4382,
+      "step": 45650
+    },
+    {
+      "epoch": 137.12,
+      "grad_norm": 11.815271377563477,
+      "learning_rate": 5.4294294294294304e-06,
+      "loss": 0.4264,
+      "step": 45660
+    },
+    {
+      "epoch": 137.15,
+      "grad_norm": 20.612194061279297,
+      "learning_rate": 5.428428428428428e-06,
+      "loss": 0.3702,
+      "step": 45670
+    },
+    {
+      "epoch": 137.18,
+      "grad_norm": 10.594101905822754,
+      "learning_rate": 5.427427427427428e-06,
+      "loss": 0.3941,
+      "step": 45680
+    },
+    {
+      "epoch": 137.21,
+      "grad_norm": 15.799530982971191,
+      "learning_rate": 5.426426426426426e-06,
+      "loss": 0.3535,
+      "step": 45690
+    },
+    {
+      "epoch": 137.24,
+      "grad_norm": 13.020283699035645,
+      "learning_rate": 5.425425425425426e-06,
+      "loss": 0.4087,
+      "step": 45700
+    },
+    {
+      "epoch": 137.27,
+      "grad_norm": 15.170722961425781,
+      "learning_rate": 5.424424424424425e-06,
+      "loss": 0.4064,
+      "step": 45710
+    },
+    {
+      "epoch": 137.3,
+      "grad_norm": 13.898157119750977,
+      "learning_rate": 5.423423423423424e-06,
+      "loss": 0.4215,
+      "step": 45720
+    },
+    {
+      "epoch": 137.33,
+      "grad_norm": 9.888854026794434,
+      "learning_rate": 5.422422422422423e-06,
+      "loss": 0.3936,
+      "step": 45730
+    },
+    {
+      "epoch": 137.36,
+      "grad_norm": 13.180124282836914,
+      "learning_rate": 5.421421421421423e-06,
+      "loss": 0.3967,
+      "step": 45740
+    },
+    {
+      "epoch": 137.39,
+      "grad_norm": 15.97644329071045,
+      "learning_rate": 5.420420420420421e-06,
+      "loss": 0.403,
+      "step": 45750
+    },
+    {
+      "epoch": 137.42,
+      "grad_norm": 13.12777328491211,
+      "learning_rate": 5.41941941941942e-06,
+      "loss": 0.3338,
+      "step": 45760
+    },
+    {
+      "epoch": 137.45,
+      "grad_norm": 16.321786880493164,
+      "learning_rate": 5.418418418418418e-06,
+      "loss": 0.4313,
+      "step": 45770
+    },
+    {
+      "epoch": 137.48,
+      "grad_norm": 15.795038223266602,
+      "learning_rate": 5.417417417417418e-06,
+      "loss": 0.3751,
+      "step": 45780
+    },
+    {
+      "epoch": 137.51,
+      "grad_norm": 14.829188346862793,
+      "learning_rate": 5.416416416416416e-06,
+      "loss": 0.391,
+      "step": 45790
+    },
+    {
+      "epoch": 137.54,
+      "grad_norm": 18.132413864135742,
+      "learning_rate": 5.415415415415416e-06,
+      "loss": 0.4735,
+      "step": 45800
+    },
+    {
+      "epoch": 137.57,
+      "grad_norm": 16.825910568237305,
+      "learning_rate": 5.414414414414415e-06,
+      "loss": 0.4259,
+      "step": 45810
+    },
+    {
+      "epoch": 137.6,
+      "grad_norm": 18.321134567260742,
+      "learning_rate": 5.413413413413414e-06,
+      "loss": 0.3736,
+      "step": 45820
+    },
+    {
+      "epoch": 137.63,
+      "grad_norm": 14.287819862365723,
+      "learning_rate": 5.412412412412413e-06,
+      "loss": 0.4322,
+      "step": 45830
+    },
+    {
+      "epoch": 137.66,
+      "grad_norm": 12.988914489746094,
+      "learning_rate": 5.411411411411411e-06,
+      "loss": 0.4194,
+      "step": 45840
+    },
+    {
+      "epoch": 137.69,
+      "grad_norm": 21.948410034179688,
+      "learning_rate": 5.4104104104104105e-06,
+      "loss": 0.3818,
+      "step": 45850
+    },
+    {
+      "epoch": 137.72,
+      "grad_norm": 12.76795768737793,
+      "learning_rate": 5.40940940940941e-06,
+      "loss": 0.3766,
+      "step": 45860
+    },
+    {
+      "epoch": 137.75,
+      "grad_norm": 17.971708297729492,
+      "learning_rate": 5.4084084084084086e-06,
+      "loss": 0.4572,
+      "step": 45870
+    },
+    {
+      "epoch": 137.78,
+      "grad_norm": 10.760985374450684,
+      "learning_rate": 5.407407407407408e-06,
+      "loss": 0.4262,
+      "step": 45880
+    },
+    {
+      "epoch": 137.81,
+      "grad_norm": 12.426521301269531,
+      "learning_rate": 5.4064064064064074e-06,
+      "loss": 0.3918,
+      "step": 45890
+    },
+    {
+      "epoch": 137.84,
+      "grad_norm": 14.156126022338867,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.3997,
+      "step": 45900
+    },
+    {
+      "epoch": 137.87,
+      "grad_norm": 14.061455726623535,
+      "learning_rate": 5.4044044044044055e-06,
+      "loss": 0.4274,
+      "step": 45910
+    },
+    {
+      "epoch": 137.9,
+      "grad_norm": 18.042293548583984,
+      "learning_rate": 5.403403403403403e-06,
+      "loss": 0.4059,
+      "step": 45920
+    },
+    {
+      "epoch": 137.93,
+      "grad_norm": 30.02056312561035,
+      "learning_rate": 5.402402402402403e-06,
+      "loss": 0.4229,
+      "step": 45930
+    },
+    {
+      "epoch": 137.96,
+      "grad_norm": 28.83627700805664,
+      "learning_rate": 5.401401401401401e-06,
+      "loss": 0.4714,
+      "step": 45940
+    },
+    {
+      "epoch": 137.99,
+      "grad_norm": 22.390695571899414,
+      "learning_rate": 5.400400400400401e-06,
+      "loss": 0.3752,
+      "step": 45950
+    },
+    {
+      "epoch": 138.0,
+      "eval_accuracy": 0.8656,
+      "eval_loss": 0.5128827691078186,
+      "eval_runtime": 12.7833,
+      "eval_samples_per_second": 782.272,
+      "eval_steps_per_second": 3.129,
+      "step": 45954
+    },
+    {
+      "epoch": 138.02,
+      "grad_norm": 15.484604835510254,
+      "learning_rate": 5.3993993993994e-06,
+      "loss": 0.3958,
+      "step": 45960
+    },
+    {
+      "epoch": 138.05,
+      "grad_norm": 15.495908737182617,
+      "learning_rate": 5.398398398398399e-06,
+      "loss": 0.4401,
+      "step": 45970
+    },
+    {
+      "epoch": 138.08,
+      "grad_norm": 15.310026168823242,
+      "learning_rate": 5.397397397397398e-06,
+      "loss": 0.4182,
+      "step": 45980
+    },
+    {
+      "epoch": 138.11,
+      "grad_norm": 17.15497398376465,
+      "learning_rate": 5.396396396396398e-06,
+      "loss": 0.4192,
+      "step": 45990
+    },
+    {
+      "epoch": 138.14,
+      "grad_norm": 17.395097732543945,
+      "learning_rate": 5.395395395395396e-06,
+      "loss": 0.4433,
+      "step": 46000
+    },
+    {
+      "epoch": 138.17,
+      "grad_norm": 23.71068572998047,
+      "learning_rate": 5.394394394394394e-06,
+      "loss": 0.3616,
+      "step": 46010
+    },
+    {
+      "epoch": 138.2,
+      "grad_norm": 20.026325225830078,
+      "learning_rate": 5.393393393393393e-06,
+      "loss": 0.4305,
+      "step": 46020
+    },
+    {
+      "epoch": 138.23,
+      "grad_norm": 20.209915161132812,
+      "learning_rate": 5.392392392392393e-06,
+      "loss": 0.4204,
+      "step": 46030
+    },
+    {
+      "epoch": 138.26,
+      "grad_norm": 13.30009937286377,
+      "learning_rate": 5.3913913913913914e-06,
+      "loss": 0.4491,
+      "step": 46040
+    },
+    {
+      "epoch": 138.29,
+      "grad_norm": 21.37310218811035,
+      "learning_rate": 5.390390390390391e-06,
+      "loss": 0.4062,
+      "step": 46050
+    },
+    {
+      "epoch": 138.32,
+      "grad_norm": 15.528406143188477,
+      "learning_rate": 5.38938938938939e-06,
+      "loss": 0.3748,
+      "step": 46060
+    },
+    {
+      "epoch": 138.35,
+      "grad_norm": 12.24921703338623,
+      "learning_rate": 5.388388388388389e-06,
+      "loss": 0.369,
+      "step": 46070
+    },
+    {
+      "epoch": 138.38,
+      "grad_norm": 16.489912033081055,
+      "learning_rate": 5.387387387387388e-06,
+      "loss": 0.4228,
+      "step": 46080
+    },
+    {
+      "epoch": 138.41,
+      "grad_norm": 21.000896453857422,
+      "learning_rate": 5.386386386386386e-06,
+      "loss": 0.3812,
+      "step": 46090
+    },
+    {
+      "epoch": 138.44,
+      "grad_norm": 12.137725830078125,
+      "learning_rate": 5.3853853853853856e-06,
+      "loss": 0.4498,
+      "step": 46100
+    },
+    {
+      "epoch": 138.47,
+      "grad_norm": 15.099352836608887,
+      "learning_rate": 5.384384384384385e-06,
+      "loss": 0.4278,
+      "step": 46110
+    },
+    {
+      "epoch": 138.5,
+      "grad_norm": 13.437173843383789,
+      "learning_rate": 5.383383383383384e-06,
+      "loss": 0.4387,
+      "step": 46120
+    },
+    {
+      "epoch": 138.53,
+      "grad_norm": 11.183767318725586,
+      "learning_rate": 5.382382382382383e-06,
+      "loss": 0.418,
+      "step": 46130
+    },
+    {
+      "epoch": 138.56,
+      "grad_norm": 17.82787322998047,
+      "learning_rate": 5.381381381381382e-06,
+      "loss": 0.4505,
+      "step": 46140
+    },
+    {
+      "epoch": 138.59,
+      "grad_norm": 17.821224212646484,
+      "learning_rate": 5.380380380380381e-06,
+      "loss": 0.4118,
+      "step": 46150
+    },
+    {
+      "epoch": 138.62,
+      "grad_norm": 15.795697212219238,
+      "learning_rate": 5.3793793793793805e-06,
+      "loss": 0.3257,
+      "step": 46160
+    },
+    {
+      "epoch": 138.65,
+      "grad_norm": 15.848118782043457,
+      "learning_rate": 5.378378378378378e-06,
+      "loss": 0.4004,
+      "step": 46170
+    },
+    {
+      "epoch": 138.68,
+      "grad_norm": 14.52973747253418,
+      "learning_rate": 5.377377377377378e-06,
+      "loss": 0.4109,
+      "step": 46180
+    },
+    {
+      "epoch": 138.71,
+      "grad_norm": 16.881526947021484,
+      "learning_rate": 5.376376376376376e-06,
+      "loss": 0.3624,
+      "step": 46190
+    },
+    {
+      "epoch": 138.74,
+      "grad_norm": 13.81805419921875,
+      "learning_rate": 5.375375375375376e-06,
+      "loss": 0.3841,
+      "step": 46200
+    },
+    {
+      "epoch": 138.77,
+      "grad_norm": 10.629168510437012,
+      "learning_rate": 5.374374374374375e-06,
+      "loss": 0.4206,
+      "step": 46210
+    },
+    {
+      "epoch": 138.8,
+      "grad_norm": 10.515238761901855,
+      "learning_rate": 5.373373373373374e-06,
+      "loss": 0.4066,
+      "step": 46220
+    },
+    {
+      "epoch": 138.83,
+      "grad_norm": 11.75940990447998,
+      "learning_rate": 5.372372372372373e-06,
+      "loss": 0.4064,
+      "step": 46230
+    },
+    {
+      "epoch": 138.86,
+      "grad_norm": 19.362245559692383,
+      "learning_rate": 5.371371371371372e-06,
+      "loss": 0.4042,
+      "step": 46240
+    },
+    {
+      "epoch": 138.89,
+      "grad_norm": 13.299408912658691,
+      "learning_rate": 5.370370370370371e-06,
+      "loss": 0.419,
+      "step": 46250
+    },
+    {
+      "epoch": 138.92,
+      "grad_norm": 18.25461196899414,
+      "learning_rate": 5.369369369369369e-06,
+      "loss": 0.4515,
+      "step": 46260
+    },
+    {
+      "epoch": 138.95,
+      "grad_norm": 22.340604782104492,
+      "learning_rate": 5.3683683683683684e-06,
+      "loss": 0.4532,
+      "step": 46270
+    },
+    {
+      "epoch": 138.98,
+      "grad_norm": 31.110719680786133,
+      "learning_rate": 5.367367367367368e-06,
+      "loss": 0.4338,
+      "step": 46280
+    },
+    {
+      "epoch": 139.0,
+      "eval_accuracy": 0.8654,
+      "eval_loss": 0.5146753191947937,
+      "eval_runtime": 12.6769,
+      "eval_samples_per_second": 788.836,
+      "eval_steps_per_second": 3.155,
+      "step": 46287
+    },
+    {
+      "epoch": 139.01,
+      "grad_norm": 15.712041854858398,
+      "learning_rate": 5.3663663663663665e-06,
+      "loss": 0.4278,
+      "step": 46290
+    },
+    {
+      "epoch": 139.04,
+      "grad_norm": 16.59680938720703,
+      "learning_rate": 5.365365365365366e-06,
+      "loss": 0.3815,
+      "step": 46300
+    },
+    {
+      "epoch": 139.07,
+      "grad_norm": 12.72437858581543,
+      "learning_rate": 5.364364364364365e-06,
+      "loss": 0.4086,
+      "step": 46310
+    },
+    {
+      "epoch": 139.1,
+      "grad_norm": 17.02048110961914,
+      "learning_rate": 5.363363363363364e-06,
+      "loss": 0.4147,
+      "step": 46320
+    },
+    {
+      "epoch": 139.13,
+      "grad_norm": 17.88872718811035,
+      "learning_rate": 5.362362362362363e-06,
+      "loss": 0.3676,
+      "step": 46330
+    },
+    {
+      "epoch": 139.16,
+      "grad_norm": 16.355745315551758,
+      "learning_rate": 5.361361361361361e-06,
+      "loss": 0.444,
+      "step": 46340
+    },
+    {
+      "epoch": 139.19,
+      "grad_norm": 17.637939453125,
+      "learning_rate": 5.360360360360361e-06,
+      "loss": 0.4052,
+      "step": 46350
+    },
+    {
+      "epoch": 139.22,
+      "grad_norm": 17.754919052124023,
+      "learning_rate": 5.359359359359359e-06,
+      "loss": 0.373,
+      "step": 46360
+    },
+    {
+      "epoch": 139.25,
+      "grad_norm": 13.906715393066406,
+      "learning_rate": 5.358358358358359e-06,
+      "loss": 0.422,
+      "step": 46370
+    },
+    {
+      "epoch": 139.28,
+      "grad_norm": 18.733137130737305,
+      "learning_rate": 5.357357357357358e-06,
+      "loss": 0.4215,
+      "step": 46380
+    },
+    {
+      "epoch": 139.31,
+      "grad_norm": 17.704591751098633,
+      "learning_rate": 5.356356356356357e-06,
+      "loss": 0.4197,
+      "step": 46390
+    },
+    {
+      "epoch": 139.34,
+      "grad_norm": 12.566475868225098,
+      "learning_rate": 5.355355355355356e-06,
+      "loss": 0.3915,
+      "step": 46400
+    },
+    {
+      "epoch": 139.37,
+      "grad_norm": 16.498661041259766,
+      "learning_rate": 5.3543543543543555e-06,
+      "loss": 0.4436,
+      "step": 46410
+    },
+    {
+      "epoch": 139.4,
+      "grad_norm": 14.391111373901367,
+      "learning_rate": 5.353353353353353e-06,
+      "loss": 0.3917,
+      "step": 46420
+    },
+    {
+      "epoch": 139.43,
+      "grad_norm": 11.935248374938965,
+      "learning_rate": 5.352352352352353e-06,
+      "loss": 0.4034,
+      "step": 46430
+    },
+    {
+      "epoch": 139.46,
+      "grad_norm": 13.272345542907715,
+      "learning_rate": 5.351351351351351e-06,
+      "loss": 0.3844,
+      "step": 46440
+    },
+    {
+      "epoch": 139.49,
+      "grad_norm": 17.152494430541992,
+      "learning_rate": 5.350350350350351e-06,
+      "loss": 0.4006,
+      "step": 46450
+    },
+    {
+      "epoch": 139.52,
+      "grad_norm": 20.245548248291016,
+      "learning_rate": 5.34934934934935e-06,
+      "loss": 0.4021,
+      "step": 46460
+    },
+    {
+      "epoch": 139.55,
+      "grad_norm": 15.919865608215332,
+      "learning_rate": 5.348348348348349e-06,
+      "loss": 0.4055,
+      "step": 46470
+    },
+    {
+      "epoch": 139.58,
+      "grad_norm": 13.206727981567383,
+      "learning_rate": 5.347347347347348e-06,
+      "loss": 0.4139,
+      "step": 46480
+    },
+    {
+      "epoch": 139.61,
+      "grad_norm": 17.116914749145508,
+      "learning_rate": 5.346346346346347e-06,
+      "loss": 0.37,
+      "step": 46490
+    },
+    {
+      "epoch": 139.64,
+      "grad_norm": 13.55545711517334,
+      "learning_rate": 5.345345345345346e-06,
+      "loss": 0.4046,
+      "step": 46500
+    },
+    {
+      "epoch": 139.67,
+      "grad_norm": 16.477855682373047,
+      "learning_rate": 5.344344344344344e-06,
+      "loss": 0.4481,
+      "step": 46510
+    },
+    {
+      "epoch": 139.7,
+      "grad_norm": 13.791865348815918,
+      "learning_rate": 5.3433433433433435e-06,
+      "loss": 0.3988,
+      "step": 46520
+    },
+    {
+      "epoch": 139.73,
+      "grad_norm": 16.43255615234375,
+      "learning_rate": 5.342342342342343e-06,
+      "loss": 0.4287,
+      "step": 46530
+    },
+    {
+      "epoch": 139.76,
+      "grad_norm": 18.719703674316406,
+      "learning_rate": 5.3413413413413415e-06,
+      "loss": 0.433,
+      "step": 46540
+    },
+    {
+      "epoch": 139.79,
+      "grad_norm": 15.71118450164795,
+      "learning_rate": 5.340340340340341e-06,
+      "loss": 0.4099,
+      "step": 46550
+    },
+    {
+      "epoch": 139.82,
+      "grad_norm": 19.908790588378906,
+      "learning_rate": 5.33933933933934e-06,
+      "loss": 0.4391,
+      "step": 46560
+    },
+    {
+      "epoch": 139.85,
+      "grad_norm": 19.294246673583984,
+      "learning_rate": 5.338338338338339e-06,
+      "loss": 0.3725,
+      "step": 46570
+    },
+    {
+      "epoch": 139.88,
+      "grad_norm": 13.39482307434082,
+      "learning_rate": 5.3373373373373384e-06,
+      "loss": 0.4434,
+      "step": 46580
+    },
+    {
+      "epoch": 139.91,
+      "grad_norm": 19.41019630432129,
+      "learning_rate": 5.336336336336336e-06,
+      "loss": 0.3317,
+      "step": 46590
+    },
+    {
+      "epoch": 139.94,
+      "grad_norm": 13.709493637084961,
+      "learning_rate": 5.335335335335336e-06,
+      "loss": 0.3989,
+      "step": 46600
+    },
+    {
+      "epoch": 139.97,
+      "grad_norm": 16.12827491760254,
+      "learning_rate": 5.334334334334334e-06,
+      "loss": 0.4328,
+      "step": 46610
+    },
+    {
+      "epoch": 140.0,
+      "grad_norm": 122.11381530761719,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.4865,
+      "step": 46620
+    },
+    {
+      "epoch": 140.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5171416997909546,
+      "eval_runtime": 12.7477,
+      "eval_samples_per_second": 784.453,
+      "eval_steps_per_second": 3.138,
+      "step": 46620
+    },
+    {
+      "epoch": 140.03,
+      "grad_norm": 15.708036422729492,
+      "learning_rate": 5.332332332332333e-06,
+      "loss": 0.3819,
+      "step": 46630
+    },
+    {
+      "epoch": 140.06,
+      "grad_norm": 17.828716278076172,
+      "learning_rate": 5.331331331331332e-06,
+      "loss": 0.4233,
+      "step": 46640
+    },
+    {
+      "epoch": 140.09,
+      "grad_norm": 17.769208908081055,
+      "learning_rate": 5.330330330330331e-06,
+      "loss": 0.3882,
+      "step": 46650
+    },
+    {
+      "epoch": 140.12,
+      "grad_norm": 19.39164924621582,
+      "learning_rate": 5.329329329329331e-06,
+      "loss": 0.4157,
+      "step": 46660
+    },
+    {
+      "epoch": 140.15,
+      "grad_norm": 18.069766998291016,
+      "learning_rate": 5.328328328328328e-06,
+      "loss": 0.3728,
+      "step": 46670
+    },
+    {
+      "epoch": 140.18,
+      "grad_norm": 17.977365493774414,
+      "learning_rate": 5.327327327327328e-06,
+      "loss": 0.3859,
+      "step": 46680
+    },
+    {
+      "epoch": 140.21,
+      "grad_norm": 17.32118797302246,
+      "learning_rate": 5.326326326326326e-06,
+      "loss": 0.4503,
+      "step": 46690
+    },
+    {
+      "epoch": 140.24,
+      "grad_norm": 11.785441398620605,
+      "learning_rate": 5.325325325325326e-06,
+      "loss": 0.4191,
+      "step": 46700
+    },
+    {
+      "epoch": 140.27,
+      "grad_norm": 21.927698135375977,
+      "learning_rate": 5.324324324324324e-06,
+      "loss": 0.4412,
+      "step": 46710
+    },
+    {
+      "epoch": 140.3,
+      "grad_norm": 13.056880950927734,
+      "learning_rate": 5.323323323323324e-06,
+      "loss": 0.3962,
+      "step": 46720
+    },
+    {
+      "epoch": 140.33,
+      "grad_norm": 20.001680374145508,
+      "learning_rate": 5.322322322322323e-06,
+      "loss": 0.4114,
+      "step": 46730
+    },
+    {
+      "epoch": 140.36,
+      "grad_norm": 21.956287384033203,
+      "learning_rate": 5.321321321321322e-06,
+      "loss": 0.3569,
+      "step": 46740
+    },
+    {
+      "epoch": 140.39,
+      "grad_norm": 24.25130844116211,
+      "learning_rate": 5.320320320320321e-06,
+      "loss": 0.3877,
+      "step": 46750
+    },
+    {
+      "epoch": 140.42,
+      "grad_norm": 11.36959171295166,
+      "learning_rate": 5.319319319319319e-06,
+      "loss": 0.3564,
+      "step": 46760
+    },
+    {
+      "epoch": 140.45,
+      "grad_norm": 20.846040725708008,
+      "learning_rate": 5.3183183183183185e-06,
+      "loss": 0.467,
+      "step": 46770
+    },
+    {
+      "epoch": 140.48,
+      "grad_norm": 15.551453590393066,
+      "learning_rate": 5.317317317317318e-06,
+      "loss": 0.4233,
+      "step": 46780
+    },
+    {
+      "epoch": 140.51,
+      "grad_norm": 21.489173889160156,
+      "learning_rate": 5.3163163163163165e-06,
+      "loss": 0.4416,
+      "step": 46790
+    },
+    {
+      "epoch": 140.54,
+      "grad_norm": 23.226818084716797,
+      "learning_rate": 5.315315315315316e-06,
+      "loss": 0.3982,
+      "step": 46800
+    },
+    {
+      "epoch": 140.57,
+      "grad_norm": 28.338438034057617,
+      "learning_rate": 5.314314314314315e-06,
+      "loss": 0.3893,
+      "step": 46810
+    },
+    {
+      "epoch": 140.6,
+      "grad_norm": 14.044163703918457,
+      "learning_rate": 5.313313313313314e-06,
+      "loss": 0.3967,
+      "step": 46820
+    },
+    {
+      "epoch": 140.63,
+      "grad_norm": 16.529617309570312,
+      "learning_rate": 5.3123123123123135e-06,
+      "loss": 0.3818,
+      "step": 46830
+    },
+    {
+      "epoch": 140.66,
+      "grad_norm": 18.703332901000977,
+      "learning_rate": 5.311311311311311e-06,
+      "loss": 0.363,
+      "step": 46840
+    },
+    {
+      "epoch": 140.69,
+      "grad_norm": 28.94742774963379,
+      "learning_rate": 5.310310310310311e-06,
+      "loss": 0.4387,
+      "step": 46850
+    },
+    {
+      "epoch": 140.72,
+      "grad_norm": 22.714984893798828,
+      "learning_rate": 5.309309309309309e-06,
+      "loss": 0.4397,
+      "step": 46860
+    },
+    {
+      "epoch": 140.75,
+      "grad_norm": 20.156341552734375,
+      "learning_rate": 5.308308308308309e-06,
+      "loss": 0.3792,
+      "step": 46870
+    },
+    {
+      "epoch": 140.78,
+      "grad_norm": 15.797107696533203,
+      "learning_rate": 5.307307307307308e-06,
+      "loss": 0.4609,
+      "step": 46880
+    },
+    {
+      "epoch": 140.81,
+      "grad_norm": 12.6548490524292,
+      "learning_rate": 5.306306306306307e-06,
+      "loss": 0.4422,
+      "step": 46890
+    },
+    {
+      "epoch": 140.84,
+      "grad_norm": 12.723206520080566,
+      "learning_rate": 5.305305305305306e-06,
+      "loss": 0.404,
+      "step": 46900
+    },
+    {
+      "epoch": 140.87,
+      "grad_norm": 16.032873153686523,
+      "learning_rate": 5.304304304304306e-06,
+      "loss": 0.429,
+      "step": 46910
+    },
+    {
+      "epoch": 140.9,
+      "grad_norm": 11.580810546875,
+      "learning_rate": 5.303303303303303e-06,
+      "loss": 0.38,
+      "step": 46920
+    },
+    {
+      "epoch": 140.93,
+      "grad_norm": 15.842686653137207,
+      "learning_rate": 5.302302302302302e-06,
+      "loss": 0.3711,
+      "step": 46930
+    },
+    {
+      "epoch": 140.96,
+      "grad_norm": 10.958782196044922,
+      "learning_rate": 5.301301301301301e-06,
+      "loss": 0.3658,
+      "step": 46940
+    },
+    {
+      "epoch": 140.99,
+      "grad_norm": 18.949390411376953,
+      "learning_rate": 5.300300300300301e-06,
+      "loss": 0.4771,
+      "step": 46950
+    },
+    {
+      "epoch": 141.0,
+      "eval_accuracy": 0.865,
+      "eval_loss": 0.5166191458702087,
+      "eval_runtime": 12.7439,
+      "eval_samples_per_second": 784.692,
+      "eval_steps_per_second": 3.139,
+      "step": 46953
+    },
+    {
+      "epoch": 141.02,
+      "grad_norm": 14.615742683410645,
+      "learning_rate": 5.2992992992992994e-06,
+      "loss": 0.4915,
+      "step": 46960
+    },
+    {
+      "epoch": 141.05,
+      "grad_norm": 15.17050838470459,
+      "learning_rate": 5.298298298298299e-06,
+      "loss": 0.4357,
+      "step": 46970
+    },
+    {
+      "epoch": 141.08,
+      "grad_norm": 11.479752540588379,
+      "learning_rate": 5.297297297297298e-06,
+      "loss": 0.4172,
+      "step": 46980
+    },
+    {
+      "epoch": 141.11,
+      "grad_norm": 27.362213134765625,
+      "learning_rate": 5.296296296296297e-06,
+      "loss": 0.3991,
+      "step": 46990
+    },
+    {
+      "epoch": 141.14,
+      "grad_norm": 24.178678512573242,
+      "learning_rate": 5.2952952952952955e-06,
+      "loss": 0.4626,
+      "step": 47000
+    },
+    {
+      "epoch": 141.17,
+      "grad_norm": 12.786059379577637,
+      "learning_rate": 5.294294294294294e-06,
+      "loss": 0.3743,
+      "step": 47010
+    },
+    {
+      "epoch": 141.2,
+      "grad_norm": 13.822656631469727,
+      "learning_rate": 5.2932932932932935e-06,
+      "loss": 0.4194,
+      "step": 47020
+    },
+    {
+      "epoch": 141.23,
+      "grad_norm": 11.318143844604492,
+      "learning_rate": 5.292292292292293e-06,
+      "loss": 0.4107,
+      "step": 47030
+    },
+    {
+      "epoch": 141.26,
+      "grad_norm": 11.903264999389648,
+      "learning_rate": 5.291291291291292e-06,
+      "loss": 0.4334,
+      "step": 47040
+    },
+    {
+      "epoch": 141.29,
+      "grad_norm": 17.204153060913086,
+      "learning_rate": 5.290290290290291e-06,
+      "loss": 0.4124,
+      "step": 47050
+    },
+    {
+      "epoch": 141.32,
+      "grad_norm": 14.175749778747559,
+      "learning_rate": 5.28928928928929e-06,
+      "loss": 0.3173,
+      "step": 47060
+    },
+    {
+      "epoch": 141.35,
+      "grad_norm": 16.048776626586914,
+      "learning_rate": 5.288288288288289e-06,
+      "loss": 0.4006,
+      "step": 47070
+    },
+    {
+      "epoch": 141.38,
+      "grad_norm": 15.027162551879883,
+      "learning_rate": 5.2872872872872885e-06,
+      "loss": 0.413,
+      "step": 47080
+    },
+    {
+      "epoch": 141.41,
+      "grad_norm": 15.729867935180664,
+      "learning_rate": 5.286286286286286e-06,
+      "loss": 0.383,
+      "step": 47090
+    },
+    {
+      "epoch": 141.44,
+      "grad_norm": 24.54653549194336,
+      "learning_rate": 5.285285285285286e-06,
+      "loss": 0.4501,
+      "step": 47100
+    },
+    {
+      "epoch": 141.47,
+      "grad_norm": 16.161144256591797,
+      "learning_rate": 5.284284284284284e-06,
+      "loss": 0.3916,
+      "step": 47110
+    },
+    {
+      "epoch": 141.5,
+      "grad_norm": 13.642569541931152,
+      "learning_rate": 5.283283283283284e-06,
+      "loss": 0.4332,
+      "step": 47120
+    },
+    {
+      "epoch": 141.53,
+      "grad_norm": 19.74559211730957,
+      "learning_rate": 5.282282282282283e-06,
+      "loss": 0.3722,
+      "step": 47130
+    },
+    {
+      "epoch": 141.56,
+      "grad_norm": 16.259286880493164,
+      "learning_rate": 5.281281281281282e-06,
+      "loss": 0.4064,
+      "step": 47140
+    },
+    {
+      "epoch": 141.59,
+      "grad_norm": 10.896443367004395,
+      "learning_rate": 5.280280280280281e-06,
+      "loss": 0.4082,
+      "step": 47150
+    },
+    {
+      "epoch": 141.62,
+      "grad_norm": 21.950929641723633,
+      "learning_rate": 5.27927927927928e-06,
+      "loss": 0.4136,
+      "step": 47160
+    },
+    {
+      "epoch": 141.65,
+      "grad_norm": 14.783012390136719,
+      "learning_rate": 5.278278278278278e-06,
+      "loss": 0.3873,
+      "step": 47170
+    },
+    {
+      "epoch": 141.68,
+      "grad_norm": 18.013639450073242,
+      "learning_rate": 5.277277277277277e-06,
+      "loss": 0.39,
+      "step": 47180
+    },
+    {
+      "epoch": 141.71,
+      "grad_norm": 13.128569602966309,
+      "learning_rate": 5.2762762762762764e-06,
+      "loss": 0.4066,
+      "step": 47190
+    },
+    {
+      "epoch": 141.74,
+      "grad_norm": 13.49650764465332,
+      "learning_rate": 5.275275275275276e-06,
+      "loss": 0.3848,
+      "step": 47200
+    },
+    {
+      "epoch": 141.77,
+      "grad_norm": 15.356809616088867,
+      "learning_rate": 5.2742742742742745e-06,
+      "loss": 0.4131,
+      "step": 47210
+    },
+    {
+      "epoch": 141.8,
+      "grad_norm": 19.70092010498047,
+      "learning_rate": 5.273273273273274e-06,
+      "loss": 0.4355,
+      "step": 47220
+    },
+    {
+      "epoch": 141.83,
+      "grad_norm": 16.939645767211914,
+      "learning_rate": 5.272272272272273e-06,
+      "loss": 0.3663,
+      "step": 47230
+    },
+    {
+      "epoch": 141.86,
+      "grad_norm": 18.106853485107422,
+      "learning_rate": 5.271271271271272e-06,
+      "loss": 0.4223,
+      "step": 47240
+    },
+    {
+      "epoch": 141.89,
+      "grad_norm": 14.965764045715332,
+      "learning_rate": 5.2702702702702705e-06,
+      "loss": 0.3575,
+      "step": 47250
+    },
+    {
+      "epoch": 141.92,
+      "grad_norm": 16.336091995239258,
+      "learning_rate": 5.269269269269269e-06,
+      "loss": 0.4064,
+      "step": 47260
+    },
+    {
+      "epoch": 141.95,
+      "grad_norm": 29.80402374267578,
+      "learning_rate": 5.268268268268269e-06,
+      "loss": 0.4173,
+      "step": 47270
+    },
+    {
+      "epoch": 141.98,
+      "grad_norm": 14.896234512329102,
+      "learning_rate": 5.267267267267267e-06,
+      "loss": 0.4824,
+      "step": 47280
+    },
+    {
+      "epoch": 142.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5208633542060852,
+      "eval_runtime": 12.85,
+      "eval_samples_per_second": 778.213,
+      "eval_steps_per_second": 3.113,
+      "step": 47286
+    },
+    {
+      "epoch": 142.01,
+      "grad_norm": 19.523509979248047,
+      "learning_rate": 5.266266266266267e-06,
+      "loss": 0.3621,
+      "step": 47290
+    },
+    {
+      "epoch": 142.04,
+      "grad_norm": 13.22549057006836,
+      "learning_rate": 5.265265265265266e-06,
+      "loss": 0.3799,
+      "step": 47300
+    },
+    {
+      "epoch": 142.07,
+      "grad_norm": 18.622882843017578,
+      "learning_rate": 5.264264264264265e-06,
+      "loss": 0.4892,
+      "step": 47310
+    },
+    {
+      "epoch": 142.1,
+      "grad_norm": 18.625375747680664,
+      "learning_rate": 5.263263263263264e-06,
+      "loss": 0.427,
+      "step": 47320
+    },
+    {
+      "epoch": 142.13,
+      "grad_norm": 12.009258270263672,
+      "learning_rate": 5.2622622622622635e-06,
+      "loss": 0.3622,
+      "step": 47330
+    },
+    {
+      "epoch": 142.16,
+      "grad_norm": 14.679965019226074,
+      "learning_rate": 5.261261261261261e-06,
+      "loss": 0.4181,
+      "step": 47340
+    },
+    {
+      "epoch": 142.19,
+      "grad_norm": 16.904747009277344,
+      "learning_rate": 5.260260260260261e-06,
+      "loss": 0.4082,
+      "step": 47350
+    },
+    {
+      "epoch": 142.22,
+      "grad_norm": 10.413802146911621,
+      "learning_rate": 5.259259259259259e-06,
+      "loss": 0.41,
+      "step": 47360
+    },
+    {
+      "epoch": 142.25,
+      "grad_norm": 20.21839714050293,
+      "learning_rate": 5.258258258258259e-06,
+      "loss": 0.364,
+      "step": 47370
+    },
+    {
+      "epoch": 142.28,
+      "grad_norm": 14.145240783691406,
+      "learning_rate": 5.257257257257257e-06,
+      "loss": 0.3813,
+      "step": 47380
+    },
+    {
+      "epoch": 142.31,
+      "grad_norm": 17.874467849731445,
+      "learning_rate": 5.256256256256257e-06,
+      "loss": 0.3861,
+      "step": 47390
+    },
+    {
+      "epoch": 142.34,
+      "grad_norm": 14.248611450195312,
+      "learning_rate": 5.255255255255256e-06,
+      "loss": 0.4173,
+      "step": 47400
+    },
+    {
+      "epoch": 142.37,
+      "grad_norm": 18.01908302307129,
+      "learning_rate": 5.254254254254255e-06,
+      "loss": 0.413,
+      "step": 47410
+    },
+    {
+      "epoch": 142.4,
+      "grad_norm": 14.34422779083252,
+      "learning_rate": 5.2532532532532534e-06,
+      "loss": 0.4285,
+      "step": 47420
+    },
+    {
+      "epoch": 142.43,
+      "grad_norm": 11.718576431274414,
+      "learning_rate": 5.252252252252252e-06,
+      "loss": 0.3842,
+      "step": 47430
+    },
+    {
+      "epoch": 142.46,
+      "grad_norm": 11.27565860748291,
+      "learning_rate": 5.2512512512512515e-06,
+      "loss": 0.3797,
+      "step": 47440
+    },
+    {
+      "epoch": 142.49,
+      "grad_norm": 14.868589401245117,
+      "learning_rate": 5.250250250250251e-06,
+      "loss": 0.4033,
+      "step": 47450
+    },
+    {
+      "epoch": 142.52,
+      "grad_norm": 12.609210014343262,
+      "learning_rate": 5.2492492492492495e-06,
+      "loss": 0.3116,
+      "step": 47460
+    },
+    {
+      "epoch": 142.55,
+      "grad_norm": 13.401498794555664,
+      "learning_rate": 5.248248248248249e-06,
+      "loss": 0.4021,
+      "step": 47470
+    },
+    {
+      "epoch": 142.58,
+      "grad_norm": 40.36709213256836,
+      "learning_rate": 5.247247247247248e-06,
+      "loss": 0.4253,
+      "step": 47480
+    },
+    {
+      "epoch": 142.61,
+      "grad_norm": 19.081806182861328,
+      "learning_rate": 5.246246246246247e-06,
+      "loss": 0.43,
+      "step": 47490
+    },
+    {
+      "epoch": 142.64,
+      "grad_norm": 22.122793197631836,
+      "learning_rate": 5.245245245245245e-06,
+      "loss": 0.4188,
+      "step": 47500
+    },
+    {
+      "epoch": 142.67,
+      "grad_norm": 18.731855392456055,
+      "learning_rate": 5.244244244244244e-06,
+      "loss": 0.3898,
+      "step": 47510
+    },
+    {
+      "epoch": 142.7,
+      "grad_norm": 16.26125717163086,
+      "learning_rate": 5.243243243243244e-06,
+      "loss": 0.4149,
+      "step": 47520
+    },
+    {
+      "epoch": 142.73,
+      "grad_norm": 11.771873474121094,
+      "learning_rate": 5.242242242242242e-06,
+      "loss": 0.3841,
+      "step": 47530
+    },
+    {
+      "epoch": 142.76,
+      "grad_norm": 11.177895545959473,
+      "learning_rate": 5.241241241241242e-06,
+      "loss": 0.3646,
+      "step": 47540
+    },
+    {
+      "epoch": 142.79,
+      "grad_norm": 21.768701553344727,
+      "learning_rate": 5.240240240240241e-06,
+      "loss": 0.3986,
+      "step": 47550
+    },
+    {
+      "epoch": 142.82,
+      "grad_norm": 8.852099418640137,
+      "learning_rate": 5.23923923923924e-06,
+      "loss": 0.3917,
+      "step": 47560
+    },
+    {
+      "epoch": 142.85,
+      "grad_norm": 21.22300910949707,
+      "learning_rate": 5.238238238238239e-06,
+      "loss": 0.382,
+      "step": 47570
+    },
+    {
+      "epoch": 142.88,
+      "grad_norm": 19.177677154541016,
+      "learning_rate": 5.2372372372372386e-06,
+      "loss": 0.4343,
+      "step": 47580
+    },
+    {
+      "epoch": 142.91,
+      "grad_norm": 18.586254119873047,
+      "learning_rate": 5.236236236236236e-06,
+      "loss": 0.4048,
+      "step": 47590
+    },
+    {
+      "epoch": 142.94,
+      "grad_norm": 21.305130004882812,
+      "learning_rate": 5.235235235235236e-06,
+      "loss": 0.3802,
+      "step": 47600
+    },
+    {
+      "epoch": 142.97,
+      "grad_norm": 12.370361328125,
+      "learning_rate": 5.234234234234234e-06,
+      "loss": 0.4026,
+      "step": 47610
+    },
+    {
+      "epoch": 143.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5265568494796753,
+      "eval_runtime": 12.6504,
+      "eval_samples_per_second": 790.492,
+      "eval_steps_per_second": 3.162,
+      "step": 47619
+    },
+    {
+      "epoch": 143.0,
+      "grad_norm": 13.037696838378906,
+      "learning_rate": 5.233233233233234e-06,
+      "loss": 0.3951,
+      "step": 47620
+    },
+    {
+      "epoch": 143.03,
+      "grad_norm": 19.028118133544922,
+      "learning_rate": 5.232232232232232e-06,
+      "loss": 0.3881,
+      "step": 47630
+    },
+    {
+      "epoch": 143.06,
+      "grad_norm": 14.415146827697754,
+      "learning_rate": 5.231231231231232e-06,
+      "loss": 0.4068,
+      "step": 47640
+    },
+    {
+      "epoch": 143.09,
+      "grad_norm": 16.075504302978516,
+      "learning_rate": 5.230230230230231e-06,
+      "loss": 0.4106,
+      "step": 47650
+    },
+    {
+      "epoch": 143.12,
+      "grad_norm": 14.925749778747559,
+      "learning_rate": 5.22922922922923e-06,
+      "loss": 0.4085,
+      "step": 47660
+    },
+    {
+      "epoch": 143.15,
+      "grad_norm": 14.211409568786621,
+      "learning_rate": 5.2282282282282285e-06,
+      "loss": 0.3958,
+      "step": 47670
+    },
+    {
+      "epoch": 143.18,
+      "grad_norm": 8.77979850769043,
+      "learning_rate": 5.227227227227227e-06,
+      "loss": 0.3675,
+      "step": 47680
+    },
+    {
+      "epoch": 143.21,
+      "grad_norm": 17.107769012451172,
+      "learning_rate": 5.2262262262262265e-06,
+      "loss": 0.3806,
+      "step": 47690
+    },
+    {
+      "epoch": 143.24,
+      "grad_norm": 14.428203582763672,
+      "learning_rate": 5.225225225225226e-06,
+      "loss": 0.3959,
+      "step": 47700
+    },
+    {
+      "epoch": 143.27,
+      "grad_norm": 11.070428848266602,
+      "learning_rate": 5.2242242242242245e-06,
+      "loss": 0.3788,
+      "step": 47710
+    },
+    {
+      "epoch": 143.3,
+      "grad_norm": 18.9853515625,
+      "learning_rate": 5.223223223223224e-06,
+      "loss": 0.4182,
+      "step": 47720
+    },
+    {
+      "epoch": 143.33,
+      "grad_norm": 18.9548282623291,
+      "learning_rate": 5.2222222222222226e-06,
+      "loss": 0.4534,
+      "step": 47730
+    },
+    {
+      "epoch": 143.36,
+      "grad_norm": 17.235857009887695,
+      "learning_rate": 5.221221221221222e-06,
+      "loss": 0.355,
+      "step": 47740
+    },
+    {
+      "epoch": 143.39,
+      "grad_norm": 19.859477996826172,
+      "learning_rate": 5.22022022022022e-06,
+      "loss": 0.3973,
+      "step": 47750
+    },
+    {
+      "epoch": 143.42,
+      "grad_norm": 15.13555908203125,
+      "learning_rate": 5.219219219219219e-06,
+      "loss": 0.3743,
+      "step": 47760
+    },
+    {
+      "epoch": 143.45,
+      "grad_norm": 13.66828727722168,
+      "learning_rate": 5.218218218218219e-06,
+      "loss": 0.3878,
+      "step": 47770
+    },
+    {
+      "epoch": 143.48,
+      "grad_norm": 14.685361862182617,
+      "learning_rate": 5.217217217217217e-06,
+      "loss": 0.4172,
+      "step": 47780
+    },
+    {
+      "epoch": 143.51,
+      "grad_norm": 16.566986083984375,
+      "learning_rate": 5.216216216216217e-06,
+      "loss": 0.393,
+      "step": 47790
+    },
+    {
+      "epoch": 143.54,
+      "grad_norm": 19.65924072265625,
+      "learning_rate": 5.215215215215216e-06,
+      "loss": 0.3607,
+      "step": 47800
+    },
+    {
+      "epoch": 143.57,
+      "grad_norm": 12.222298622131348,
+      "learning_rate": 5.214214214214215e-06,
+      "loss": 0.4447,
+      "step": 47810
+    },
+    {
+      "epoch": 143.6,
+      "grad_norm": 10.682478904724121,
+      "learning_rate": 5.213213213213214e-06,
+      "loss": 0.4118,
+      "step": 47820
+    },
+    {
+      "epoch": 143.63,
+      "grad_norm": 21.63252830505371,
+      "learning_rate": 5.212212212212214e-06,
+      "loss": 0.391,
+      "step": 47830
+    },
+    {
+      "epoch": 143.66,
+      "grad_norm": 10.757732391357422,
+      "learning_rate": 5.211211211211211e-06,
+      "loss": 0.3769,
+      "step": 47840
+    },
+    {
+      "epoch": 143.69,
+      "grad_norm": 12.382858276367188,
+      "learning_rate": 5.21021021021021e-06,
+      "loss": 0.3673,
+      "step": 47850
+    },
+    {
+      "epoch": 143.72,
+      "grad_norm": 15.601813316345215,
+      "learning_rate": 5.209209209209209e-06,
+      "loss": 0.365,
+      "step": 47860
+    },
+    {
+      "epoch": 143.75,
+      "grad_norm": 19.92235565185547,
+      "learning_rate": 5.208208208208209e-06,
+      "loss": 0.3734,
+      "step": 47870
+    },
+    {
+      "epoch": 143.78,
+      "grad_norm": 14.979222297668457,
+      "learning_rate": 5.2072072072072074e-06,
+      "loss": 0.3684,
+      "step": 47880
+    },
+    {
+      "epoch": 143.81,
+      "grad_norm": 16.00888442993164,
+      "learning_rate": 5.206206206206207e-06,
+      "loss": 0.4077,
+      "step": 47890
+    },
+    {
+      "epoch": 143.84,
+      "grad_norm": 17.720592498779297,
+      "learning_rate": 5.205205205205206e-06,
+      "loss": 0.4405,
+      "step": 47900
+    },
+    {
+      "epoch": 143.87,
+      "grad_norm": 14.020065307617188,
+      "learning_rate": 5.204204204204205e-06,
+      "loss": 0.3898,
+      "step": 47910
+    },
+    {
+      "epoch": 143.9,
+      "grad_norm": 10.385543823242188,
+      "learning_rate": 5.2032032032032035e-06,
+      "loss": 0.3671,
+      "step": 47920
+    },
+    {
+      "epoch": 143.93,
+      "grad_norm": 14.212424278259277,
+      "learning_rate": 5.202202202202202e-06,
+      "loss": 0.3486,
+      "step": 47930
+    },
+    {
+      "epoch": 143.96,
+      "grad_norm": 16.516996383666992,
+      "learning_rate": 5.2012012012012015e-06,
+      "loss": 0.4408,
+      "step": 47940
+    },
+    {
+      "epoch": 143.99,
+      "grad_norm": 10.240898132324219,
+      "learning_rate": 5.200200200200201e-06,
+      "loss": 0.436,
+      "step": 47950
+    },
+    {
+      "epoch": 144.0,
+      "eval_accuracy": 0.8658,
+      "eval_loss": 0.5199581384658813,
+      "eval_runtime": 12.8025,
+      "eval_samples_per_second": 781.095,
+      "eval_steps_per_second": 3.124,
+      "step": 47952
+    },
+    {
+      "epoch": 144.02,
+      "grad_norm": 13.2849702835083,
+      "learning_rate": 5.1991991991991996e-06,
+      "loss": 0.3658,
+      "step": 47960
+    },
+    {
+      "epoch": 144.05,
+      "grad_norm": 12.01689338684082,
+      "learning_rate": 5.198198198198199e-06,
+      "loss": 0.419,
+      "step": 47970
+    },
+    {
+      "epoch": 144.08,
+      "grad_norm": 10.590292930603027,
+      "learning_rate": 5.197197197197198e-06,
+      "loss": 0.3701,
+      "step": 47980
+    },
+    {
+      "epoch": 144.11,
+      "grad_norm": 16.4504337310791,
+      "learning_rate": 5.196196196196197e-06,
+      "loss": 0.4459,
+      "step": 47990
+    },
+    {
+      "epoch": 144.14,
+      "grad_norm": 10.621485710144043,
+      "learning_rate": 5.195195195195195e-06,
+      "loss": 0.4104,
+      "step": 48000
+    },
+    {
+      "epoch": 144.17,
+      "grad_norm": 17.478748321533203,
+      "learning_rate": 5.194194194194194e-06,
+      "loss": 0.4186,
+      "step": 48010
+    },
+    {
+      "epoch": 144.2,
+      "grad_norm": 14.16486644744873,
+      "learning_rate": 5.193193193193194e-06,
+      "loss": 0.3974,
+      "step": 48020
+    },
+    {
+      "epoch": 144.23,
+      "grad_norm": 15.325855255126953,
+      "learning_rate": 5.192192192192192e-06,
+      "loss": 0.3905,
+      "step": 48030
+    },
+    {
+      "epoch": 144.26,
+      "grad_norm": 12.836835861206055,
+      "learning_rate": 5.191191191191192e-06,
+      "loss": 0.4724,
+      "step": 48040
+    },
+    {
+      "epoch": 144.29,
+      "grad_norm": 17.432920455932617,
+      "learning_rate": 5.190190190190191e-06,
+      "loss": 0.3844,
+      "step": 48050
+    },
+    {
+      "epoch": 144.32,
+      "grad_norm": 11.059144020080566,
+      "learning_rate": 5.18918918918919e-06,
+      "loss": 0.3841,
+      "step": 48060
+    },
+    {
+      "epoch": 144.35,
+      "grad_norm": 15.778397560119629,
+      "learning_rate": 5.188188188188189e-06,
+      "loss": 0.4076,
+      "step": 48070
+    },
+    {
+      "epoch": 144.38,
+      "grad_norm": 16.848190307617188,
+      "learning_rate": 5.187187187187187e-06,
+      "loss": 0.3559,
+      "step": 48080
+    },
+    {
+      "epoch": 144.41,
+      "grad_norm": 17.21855354309082,
+      "learning_rate": 5.186186186186186e-06,
+      "loss": 0.3694,
+      "step": 48090
+    },
+    {
+      "epoch": 144.44,
+      "grad_norm": 16.369829177856445,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.454,
+      "step": 48100
+    },
+    {
+      "epoch": 144.47,
+      "grad_norm": 19.073320388793945,
+      "learning_rate": 5.1841841841841844e-06,
+      "loss": 0.4342,
+      "step": 48110
+    },
+    {
+      "epoch": 144.5,
+      "grad_norm": 12.982419967651367,
+      "learning_rate": 5.183183183183184e-06,
+      "loss": 0.3896,
+      "step": 48120
+    },
+    {
+      "epoch": 144.53,
+      "grad_norm": 19.646404266357422,
+      "learning_rate": 5.1821821821821825e-06,
+      "loss": 0.4025,
+      "step": 48130
+    },
+    {
+      "epoch": 144.56,
+      "grad_norm": 22.33662986755371,
+      "learning_rate": 5.181181181181182e-06,
+      "loss": 0.4904,
+      "step": 48140
+    },
+    {
+      "epoch": 144.59,
+      "grad_norm": 11.163668632507324,
+      "learning_rate": 5.180180180180181e-06,
+      "loss": 0.4525,
+      "step": 48150
+    },
+    {
+      "epoch": 144.62,
+      "grad_norm": 16.271827697753906,
+      "learning_rate": 5.17917917917918e-06,
+      "loss": 0.3648,
+      "step": 48160
+    },
+    {
+      "epoch": 144.65,
+      "grad_norm": 11.610713958740234,
+      "learning_rate": 5.1781781781781785e-06,
+      "loss": 0.4342,
+      "step": 48170
+    },
+    {
+      "epoch": 144.68,
+      "grad_norm": 12.85755729675293,
+      "learning_rate": 5.177177177177177e-06,
+      "loss": 0.4158,
+      "step": 48180
+    },
+    {
+      "epoch": 144.71,
+      "grad_norm": 13.181352615356445,
+      "learning_rate": 5.1761761761761766e-06,
+      "loss": 0.468,
+      "step": 48190
+    },
+    {
+      "epoch": 144.74,
+      "grad_norm": 18.331987380981445,
+      "learning_rate": 5.175175175175175e-06,
+      "loss": 0.3873,
+      "step": 48200
+    },
+    {
+      "epoch": 144.77,
+      "grad_norm": 21.690942764282227,
+      "learning_rate": 5.174174174174175e-06,
+      "loss": 0.4328,
+      "step": 48210
+    },
+    {
+      "epoch": 144.8,
+      "grad_norm": 20.502044677734375,
+      "learning_rate": 5.173173173173174e-06,
+      "loss": 0.4036,
+      "step": 48220
+    },
+    {
+      "epoch": 144.83,
+      "grad_norm": 18.06813621520996,
+      "learning_rate": 5.172172172172173e-06,
+      "loss": 0.4084,
+      "step": 48230
+    },
+    {
+      "epoch": 144.86,
+      "grad_norm": 13.357542991638184,
+      "learning_rate": 5.171171171171172e-06,
+      "loss": 0.3997,
+      "step": 48240
+    },
+    {
+      "epoch": 144.89,
+      "grad_norm": 13.180268287658691,
+      "learning_rate": 5.17017017017017e-06,
+      "loss": 0.4005,
+      "step": 48250
+    },
+    {
+      "epoch": 144.92,
+      "grad_norm": 13.54417896270752,
+      "learning_rate": 5.169169169169169e-06,
+      "loss": 0.3679,
+      "step": 48260
+    },
+    {
+      "epoch": 144.95,
+      "grad_norm": 13.672243118286133,
+      "learning_rate": 5.168168168168169e-06,
+      "loss": 0.3517,
+      "step": 48270
+    },
+    {
+      "epoch": 144.98,
+      "grad_norm": 13.773894309997559,
+      "learning_rate": 5.167167167167167e-06,
+      "loss": 0.3487,
+      "step": 48280
+    },
+    {
+      "epoch": 145.0,
+      "eval_accuracy": 0.8644,
+      "eval_loss": 0.5184524655342102,
+      "eval_runtime": 12.7337,
+      "eval_samples_per_second": 785.318,
+      "eval_steps_per_second": 3.141,
+      "step": 48285
+    },
+    {
+      "epoch": 145.02,
+      "grad_norm": 15.033893585205078,
+      "learning_rate": 5.166166166166167e-06,
+      "loss": 0.3386,
+      "step": 48290
+    },
+    {
+      "epoch": 145.05,
+      "grad_norm": 14.428360939025879,
+      "learning_rate": 5.165165165165165e-06,
+      "loss": 0.3985,
+      "step": 48300
+    },
+    {
+      "epoch": 145.08,
+      "grad_norm": 15.598589897155762,
+      "learning_rate": 5.164164164164165e-06,
+      "loss": 0.408,
+      "step": 48310
+    },
+    {
+      "epoch": 145.11,
+      "grad_norm": 16.582252502441406,
+      "learning_rate": 5.163163163163164e-06,
+      "loss": 0.4374,
+      "step": 48320
+    },
+    {
+      "epoch": 145.14,
+      "grad_norm": 16.029922485351562,
+      "learning_rate": 5.162162162162162e-06,
+      "loss": 0.4503,
+      "step": 48330
+    },
+    {
+      "epoch": 145.17,
+      "grad_norm": 16.1333065032959,
+      "learning_rate": 5.161161161161161e-06,
+      "loss": 0.4361,
+      "step": 48340
+    },
+    {
+      "epoch": 145.2,
+      "grad_norm": 13.583585739135742,
+      "learning_rate": 5.16016016016016e-06,
+      "loss": 0.3562,
+      "step": 48350
+    },
+    {
+      "epoch": 145.23,
+      "grad_norm": 21.428085327148438,
+      "learning_rate": 5.1591591591591595e-06,
+      "loss": 0.3847,
+      "step": 48360
+    },
+    {
+      "epoch": 145.26,
+      "grad_norm": 17.782073974609375,
+      "learning_rate": 5.158158158158159e-06,
+      "loss": 0.386,
+      "step": 48370
+    },
+    {
+      "epoch": 145.29,
+      "grad_norm": 12.091958045959473,
+      "learning_rate": 5.1571571571571575e-06,
+      "loss": 0.3756,
+      "step": 48380
+    },
+    {
+      "epoch": 145.32,
+      "grad_norm": 20.306283950805664,
+      "learning_rate": 5.156156156156157e-06,
+      "loss": 0.3944,
+      "step": 48390
+    },
+    {
+      "epoch": 145.35,
+      "grad_norm": 14.184479713439941,
+      "learning_rate": 5.155155155155156e-06,
+      "loss": 0.4208,
+      "step": 48400
+    },
+    {
+      "epoch": 145.38,
+      "grad_norm": 20.03951644897461,
+      "learning_rate": 5.154154154154155e-06,
+      "loss": 0.3945,
+      "step": 48410
+    },
+    {
+      "epoch": 145.41,
+      "grad_norm": 19.502450942993164,
+      "learning_rate": 5.153153153153153e-06,
+      "loss": 0.3756,
+      "step": 48420
+    },
+    {
+      "epoch": 145.44,
+      "grad_norm": 17.298006057739258,
+      "learning_rate": 5.152152152152152e-06,
+      "loss": 0.3841,
+      "step": 48430
+    },
+    {
+      "epoch": 145.47,
+      "grad_norm": 14.357793807983398,
+      "learning_rate": 5.151151151151152e-06,
+      "loss": 0.4204,
+      "step": 48440
+    },
+    {
+      "epoch": 145.5,
+      "grad_norm": 13.546245574951172,
+      "learning_rate": 5.15015015015015e-06,
+      "loss": 0.4253,
+      "step": 48450
+    },
+    {
+      "epoch": 145.53,
+      "grad_norm": 16.01154136657715,
+      "learning_rate": 5.14914914914915e-06,
+      "loss": 0.398,
+      "step": 48460
+    },
+    {
+      "epoch": 145.56,
+      "grad_norm": 17.995258331298828,
+      "learning_rate": 5.148148148148149e-06,
+      "loss": 0.3979,
+      "step": 48470
+    },
+    {
+      "epoch": 145.59,
+      "grad_norm": 18.5677547454834,
+      "learning_rate": 5.147147147147148e-06,
+      "loss": 0.4492,
+      "step": 48480
+    },
+    {
+      "epoch": 145.62,
+      "grad_norm": 20.059953689575195,
+      "learning_rate": 5.146146146146147e-06,
+      "loss": 0.4459,
+      "step": 48490
+    },
+    {
+      "epoch": 145.65,
+      "grad_norm": 23.69488525390625,
+      "learning_rate": 5.145145145145145e-06,
+      "loss": 0.3965,
+      "step": 48500
+    },
+    {
+      "epoch": 145.68,
+      "grad_norm": 19.60682487487793,
+      "learning_rate": 5.144144144144144e-06,
+      "loss": 0.4764,
+      "step": 48510
+    },
+    {
+      "epoch": 145.71,
+      "grad_norm": 12.874115943908691,
+      "learning_rate": 5.143143143143144e-06,
+      "loss": 0.4476,
+      "step": 48520
+    },
+    {
+      "epoch": 145.74,
+      "grad_norm": 14.454435348510742,
+      "learning_rate": 5.142142142142142e-06,
+      "loss": 0.3459,
+      "step": 48530
+    },
+    {
+      "epoch": 145.77,
+      "grad_norm": 13.170857429504395,
+      "learning_rate": 5.141141141141142e-06,
+      "loss": 0.3902,
+      "step": 48540
+    },
+    {
+      "epoch": 145.8,
+      "grad_norm": 13.040736198425293,
+      "learning_rate": 5.14014014014014e-06,
+      "loss": 0.3916,
+      "step": 48550
+    },
+    {
+      "epoch": 145.83,
+      "grad_norm": 14.281366348266602,
+      "learning_rate": 5.13913913913914e-06,
+      "loss": 0.4219,
+      "step": 48560
+    },
+    {
+      "epoch": 145.86,
+      "grad_norm": 18.07878875732422,
+      "learning_rate": 5.138138138138139e-06,
+      "loss": 0.3596,
+      "step": 48570
+    },
+    {
+      "epoch": 145.89,
+      "grad_norm": 23.011062622070312,
+      "learning_rate": 5.137137137137137e-06,
+      "loss": 0.413,
+      "step": 48580
+    },
+    {
+      "epoch": 145.92,
+      "grad_norm": 18.1369571685791,
+      "learning_rate": 5.1361361361361365e-06,
+      "loss": 0.41,
+      "step": 48590
+    },
+    {
+      "epoch": 145.95,
+      "grad_norm": 13.734223365783691,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.425,
+      "step": 48600
+    },
+    {
+      "epoch": 145.98,
+      "grad_norm": 18.679189682006836,
+      "learning_rate": 5.1341341341341345e-06,
+      "loss": 0.3615,
+      "step": 48610
+    },
+    {
+      "epoch": 146.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5201944708824158,
+      "eval_runtime": 12.6556,
+      "eval_samples_per_second": 790.165,
+      "eval_steps_per_second": 3.161,
+      "step": 48618
+    },
+    {
+      "epoch": 146.01,
+      "grad_norm": 19.058521270751953,
+      "learning_rate": 5.133133133133134e-06,
+      "loss": 0.2904,
+      "step": 48620
+    },
+    {
+      "epoch": 146.04,
+      "grad_norm": 12.654631614685059,
+      "learning_rate": 5.1321321321321325e-06,
+      "loss": 0.4269,
+      "step": 48630
+    },
+    {
+      "epoch": 146.07,
+      "grad_norm": 19.038599014282227,
+      "learning_rate": 5.131131131131132e-06,
+      "loss": 0.4082,
+      "step": 48640
+    },
+    {
+      "epoch": 146.1,
+      "grad_norm": 18.63401985168457,
+      "learning_rate": 5.1301301301301306e-06,
+      "loss": 0.3939,
+      "step": 48650
+    },
+    {
+      "epoch": 146.13,
+      "grad_norm": 11.133474349975586,
+      "learning_rate": 5.12912912912913e-06,
+      "loss": 0.3772,
+      "step": 48660
+    },
+    {
+      "epoch": 146.16,
+      "grad_norm": 15.23792552947998,
+      "learning_rate": 5.128128128128128e-06,
+      "loss": 0.3939,
+      "step": 48670
+    },
+    {
+      "epoch": 146.19,
+      "grad_norm": 14.116238594055176,
+      "learning_rate": 5.127127127127127e-06,
+      "loss": 0.4151,
+      "step": 48680
+    },
+    {
+      "epoch": 146.22,
+      "grad_norm": 18.366897583007812,
+      "learning_rate": 5.126126126126127e-06,
+      "loss": 0.4048,
+      "step": 48690
+    },
+    {
+      "epoch": 146.25,
+      "grad_norm": 26.075990676879883,
+      "learning_rate": 5.125125125125125e-06,
+      "loss": 0.4294,
+      "step": 48700
+    },
+    {
+      "epoch": 146.28,
+      "grad_norm": 16.5179443359375,
+      "learning_rate": 5.124124124124125e-06,
+      "loss": 0.3945,
+      "step": 48710
+    },
+    {
+      "epoch": 146.31,
+      "grad_norm": 18.197206497192383,
+      "learning_rate": 5.123123123123124e-06,
+      "loss": 0.4627,
+      "step": 48720
+    },
+    {
+      "epoch": 146.34,
+      "grad_norm": 11.993475914001465,
+      "learning_rate": 5.122122122122123e-06,
+      "loss": 0.3776,
+      "step": 48730
+    },
+    {
+      "epoch": 146.37,
+      "grad_norm": 12.061667442321777,
+      "learning_rate": 5.121121121121122e-06,
+      "loss": 0.4209,
+      "step": 48740
+    },
+    {
+      "epoch": 146.4,
+      "grad_norm": 18.060260772705078,
+      "learning_rate": 5.12012012012012e-06,
+      "loss": 0.3885,
+      "step": 48750
+    },
+    {
+      "epoch": 146.43,
+      "grad_norm": 14.879415512084961,
+      "learning_rate": 5.119119119119119e-06,
+      "loss": 0.3628,
+      "step": 48760
+    },
+    {
+      "epoch": 146.46,
+      "grad_norm": 14.949798583984375,
+      "learning_rate": 5.118118118118118e-06,
+      "loss": 0.3812,
+      "step": 48770
+    },
+    {
+      "epoch": 146.49,
+      "grad_norm": 22.21747398376465,
+      "learning_rate": 5.117117117117117e-06,
+      "loss": 0.416,
+      "step": 48780
+    },
+    {
+      "epoch": 146.52,
+      "grad_norm": 13.513352394104004,
+      "learning_rate": 5.116116116116117e-06,
+      "loss": 0.4098,
+      "step": 48790
+    },
+    {
+      "epoch": 146.55,
+      "grad_norm": 16.96160125732422,
+      "learning_rate": 5.115115115115115e-06,
+      "loss": 0.3892,
+      "step": 48800
+    },
+    {
+      "epoch": 146.58,
+      "grad_norm": 25.692827224731445,
+      "learning_rate": 5.114114114114115e-06,
+      "loss": 0.4103,
+      "step": 48810
+    },
+    {
+      "epoch": 146.61,
+      "grad_norm": 16.650863647460938,
+      "learning_rate": 5.113113113113114e-06,
+      "loss": 0.4215,
+      "step": 48820
+    },
+    {
+      "epoch": 146.64,
+      "grad_norm": 21.987117767333984,
+      "learning_rate": 5.112112112112112e-06,
+      "loss": 0.3788,
+      "step": 48830
+    },
+    {
+      "epoch": 146.67,
+      "grad_norm": 21.978548049926758,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 0.3834,
+      "step": 48840
+    },
+    {
+      "epoch": 146.7,
+      "grad_norm": 14.778863906860352,
+      "learning_rate": 5.11011011011011e-06,
+      "loss": 0.3485,
+      "step": 48850
+    },
+    {
+      "epoch": 146.73,
+      "grad_norm": 20.152788162231445,
+      "learning_rate": 5.1091091091091095e-06,
+      "loss": 0.4174,
+      "step": 48860
+    },
+    {
+      "epoch": 146.76,
+      "grad_norm": 22.15107536315918,
+      "learning_rate": 5.108108108108108e-06,
+      "loss": 0.4355,
+      "step": 48870
+    },
+    {
+      "epoch": 146.79,
+      "grad_norm": 13.937146186828613,
+      "learning_rate": 5.1071071071071076e-06,
+      "loss": 0.4381,
+      "step": 48880
+    },
+    {
+      "epoch": 146.82,
+      "grad_norm": 15.736023902893066,
+      "learning_rate": 5.106106106106107e-06,
+      "loss": 0.4724,
+      "step": 48890
+    },
+    {
+      "epoch": 146.85,
+      "grad_norm": 15.295695304870605,
+      "learning_rate": 5.105105105105106e-06,
+      "loss": 0.4092,
+      "step": 48900
+    },
+    {
+      "epoch": 146.88,
+      "grad_norm": 18.365633010864258,
+      "learning_rate": 5.104104104104105e-06,
+      "loss": 0.3775,
+      "step": 48910
+    },
+    {
+      "epoch": 146.91,
+      "grad_norm": 15.063292503356934,
+      "learning_rate": 5.103103103103103e-06,
+      "loss": 0.4316,
+      "step": 48920
+    },
+    {
+      "epoch": 146.94,
+      "grad_norm": 14.548504829406738,
+      "learning_rate": 5.102102102102102e-06,
+      "loss": 0.44,
+      "step": 48930
+    },
+    {
+      "epoch": 146.97,
+      "grad_norm": 15.24724292755127,
+      "learning_rate": 5.101101101101102e-06,
+      "loss": 0.3652,
+      "step": 48940
+    },
+    {
+      "epoch": 147.0,
+      "grad_norm": 18.21245765686035,
+      "learning_rate": 5.1001001001001e-06,
+      "loss": 0.4317,
+      "step": 48950
+    },
+    {
+      "epoch": 147.0,
+      "eval_accuracy": 0.8658,
+      "eval_loss": 0.5197685360908508,
+      "eval_runtime": 12.9889,
+      "eval_samples_per_second": 769.885,
+      "eval_steps_per_second": 3.08,
+      "step": 48951
+    },
+    {
+      "epoch": 147.03,
+      "grad_norm": 12.064241409301758,
+      "learning_rate": 5.0990990990991e-06,
+      "loss": 0.5617,
+      "step": 48960
+    },
+    {
+      "epoch": 147.06,
+      "grad_norm": 11.663165092468262,
+      "learning_rate": 5.098098098098099e-06,
+      "loss": 0.3984,
+      "step": 48970
+    },
+    {
+      "epoch": 147.09,
+      "grad_norm": 18.790752410888672,
+      "learning_rate": 5.097097097097098e-06,
+      "loss": 0.405,
+      "step": 48980
+    },
+    {
+      "epoch": 147.12,
+      "grad_norm": 13.264315605163574,
+      "learning_rate": 5.096096096096097e-06,
+      "loss": 0.3803,
+      "step": 48990
+    },
+    {
+      "epoch": 147.15,
+      "grad_norm": 19.128368377685547,
+      "learning_rate": 5.095095095095095e-06,
+      "loss": 0.3814,
+      "step": 49000
+    },
+    {
+      "epoch": 147.18,
+      "grad_norm": 13.54724407196045,
+      "learning_rate": 5.094094094094094e-06,
+      "loss": 0.4232,
+      "step": 49010
+    },
+    {
+      "epoch": 147.21,
+      "grad_norm": 17.917064666748047,
+      "learning_rate": 5.093093093093093e-06,
+      "loss": 0.4401,
+      "step": 49020
+    },
+    {
+      "epoch": 147.24,
+      "grad_norm": 9.87697696685791,
+      "learning_rate": 5.092092092092092e-06,
+      "loss": 0.3857,
+      "step": 49030
+    },
+    {
+      "epoch": 147.27,
+      "grad_norm": 12.847880363464355,
+      "learning_rate": 5.091091091091092e-06,
+      "loss": 0.4385,
+      "step": 49040
+    },
+    {
+      "epoch": 147.3,
+      "grad_norm": 14.916487693786621,
+      "learning_rate": 5.0900900900900905e-06,
+      "loss": 0.4523,
+      "step": 49050
+    },
+    {
+      "epoch": 147.33,
+      "grad_norm": 20.728008270263672,
+      "learning_rate": 5.08908908908909e-06,
+      "loss": 0.3826,
+      "step": 49060
+    },
+    {
+      "epoch": 147.36,
+      "grad_norm": 15.296022415161133,
+      "learning_rate": 5.088088088088089e-06,
+      "loss": 0.4182,
+      "step": 49070
+    },
+    {
+      "epoch": 147.39,
+      "grad_norm": 11.604424476623535,
+      "learning_rate": 5.087087087087087e-06,
+      "loss": 0.3904,
+      "step": 49080
+    },
+    {
+      "epoch": 147.42,
+      "grad_norm": 9.612552642822266,
+      "learning_rate": 5.0860860860860865e-06,
+      "loss": 0.4149,
+      "step": 49090
+    },
+    {
+      "epoch": 147.45,
+      "grad_norm": 17.788732528686523,
+      "learning_rate": 5.085085085085085e-06,
+      "loss": 0.4214,
+      "step": 49100
+    },
+    {
+      "epoch": 147.48,
+      "grad_norm": 14.393099784851074,
+      "learning_rate": 5.0840840840840846e-06,
+      "loss": 0.4027,
+      "step": 49110
+    },
+    {
+      "epoch": 147.51,
+      "grad_norm": 14.486730575561523,
+      "learning_rate": 5.083083083083083e-06,
+      "loss": 0.343,
+      "step": 49120
+    },
+    {
+      "epoch": 147.54,
+      "grad_norm": 21.30130386352539,
+      "learning_rate": 5.082082082082083e-06,
+      "loss": 0.3462,
+      "step": 49130
+    },
+    {
+      "epoch": 147.57,
+      "grad_norm": 18.47138214111328,
+      "learning_rate": 5.081081081081082e-06,
+      "loss": 0.4061,
+      "step": 49140
+    },
+    {
+      "epoch": 147.6,
+      "grad_norm": 14.157623291015625,
+      "learning_rate": 5.080080080080081e-06,
+      "loss": 0.4234,
+      "step": 49150
+    },
+    {
+      "epoch": 147.63,
+      "grad_norm": 16.981130599975586,
+      "learning_rate": 5.07907907907908e-06,
+      "loss": 0.3652,
+      "step": 49160
+    },
+    {
+      "epoch": 147.66,
+      "grad_norm": 15.186668395996094,
+      "learning_rate": 5.078078078078078e-06,
+      "loss": 0.3972,
+      "step": 49170
+    },
+    {
+      "epoch": 147.69,
+      "grad_norm": 21.597007751464844,
+      "learning_rate": 5.077077077077077e-06,
+      "loss": 0.4253,
+      "step": 49180
+    },
+    {
+      "epoch": 147.72,
+      "grad_norm": 11.672304153442383,
+      "learning_rate": 5.076076076076077e-06,
+      "loss": 0.4166,
+      "step": 49190
+    },
+    {
+      "epoch": 147.75,
+      "grad_norm": 16.93379020690918,
+      "learning_rate": 5.075075075075075e-06,
+      "loss": 0.385,
+      "step": 49200
+    },
+    {
+      "epoch": 147.78,
+      "grad_norm": 15.99593448638916,
+      "learning_rate": 5.074074074074075e-06,
+      "loss": 0.4285,
+      "step": 49210
+    },
+    {
+      "epoch": 147.81,
+      "grad_norm": 12.933314323425293,
+      "learning_rate": 5.073073073073073e-06,
+      "loss": 0.4296,
+      "step": 49220
+    },
+    {
+      "epoch": 147.84,
+      "grad_norm": 12.5288667678833,
+      "learning_rate": 5.072072072072073e-06,
+      "loss": 0.4001,
+      "step": 49230
+    },
+    {
+      "epoch": 147.87,
+      "grad_norm": 13.52712345123291,
+      "learning_rate": 5.071071071071072e-06,
+      "loss": 0.4197,
+      "step": 49240
+    },
+    {
+      "epoch": 147.9,
+      "grad_norm": 19.138248443603516,
+      "learning_rate": 5.07007007007007e-06,
+      "loss": 0.3756,
+      "step": 49250
+    },
+    {
+      "epoch": 147.93,
+      "grad_norm": 16.074750900268555,
+      "learning_rate": 5.069069069069069e-06,
+      "loss": 0.4347,
+      "step": 49260
+    },
+    {
+      "epoch": 147.96,
+      "grad_norm": 13.791629791259766,
+      "learning_rate": 5.068068068068068e-06,
+      "loss": 0.4306,
+      "step": 49270
+    },
+    {
+      "epoch": 147.99,
+      "grad_norm": 14.313682556152344,
+      "learning_rate": 5.0670670670670675e-06,
+      "loss": 0.4171,
+      "step": 49280
+    },
+    {
+      "epoch": 148.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5236039161682129,
+      "eval_runtime": 12.5167,
+      "eval_samples_per_second": 798.935,
+      "eval_steps_per_second": 3.196,
+      "step": 49284
+    },
+    {
+      "epoch": 148.02,
+      "grad_norm": 10.859739303588867,
+      "learning_rate": 5.066066066066067e-06,
+      "loss": 0.3701,
+      "step": 49290
+    },
+    {
+      "epoch": 148.05,
+      "grad_norm": 16.46840476989746,
+      "learning_rate": 5.0650650650650655e-06,
+      "loss": 0.3615,
+      "step": 49300
+    },
+    {
+      "epoch": 148.08,
+      "grad_norm": 15.317222595214844,
+      "learning_rate": 5.064064064064065e-06,
+      "loss": 0.3943,
+      "step": 49310
+    },
+    {
+      "epoch": 148.11,
+      "grad_norm": 11.672663688659668,
+      "learning_rate": 5.0630630630630635e-06,
+      "loss": 0.3666,
+      "step": 49320
+    },
+    {
+      "epoch": 148.14,
+      "grad_norm": 15.528824806213379,
+      "learning_rate": 5.062062062062062e-06,
+      "loss": 0.36,
+      "step": 49330
+    },
+    {
+      "epoch": 148.17,
+      "grad_norm": 23.446001052856445,
+      "learning_rate": 5.061061061061061e-06,
+      "loss": 0.4644,
+      "step": 49340
+    },
+    {
+      "epoch": 148.2,
+      "grad_norm": 16.813072204589844,
+      "learning_rate": 5.06006006006006e-06,
+      "loss": 0.4406,
+      "step": 49350
+    },
+    {
+      "epoch": 148.23,
+      "grad_norm": 20.623188018798828,
+      "learning_rate": 5.05905905905906e-06,
+      "loss": 0.3791,
+      "step": 49360
+    },
+    {
+      "epoch": 148.26,
+      "grad_norm": 25.901926040649414,
+      "learning_rate": 5.058058058058058e-06,
+      "loss": 0.4282,
+      "step": 49370
+    },
+    {
+      "epoch": 148.29,
+      "grad_norm": 18.77366065979004,
+      "learning_rate": 5.057057057057058e-06,
+      "loss": 0.3967,
+      "step": 49380
+    },
+    {
+      "epoch": 148.32,
+      "grad_norm": 11.8654203414917,
+      "learning_rate": 5.056056056056057e-06,
+      "loss": 0.3983,
+      "step": 49390
+    },
+    {
+      "epoch": 148.35,
+      "grad_norm": 18.321910858154297,
+      "learning_rate": 5.055055055055056e-06,
+      "loss": 0.3817,
+      "step": 49400
+    },
+    {
+      "epoch": 148.38,
+      "grad_norm": 19.646575927734375,
+      "learning_rate": 5.054054054054054e-06,
+      "loss": 0.4237,
+      "step": 49410
+    },
+    {
+      "epoch": 148.41,
+      "grad_norm": 18.292753219604492,
+      "learning_rate": 5.053053053053053e-06,
+      "loss": 0.3874,
+      "step": 49420
+    },
+    {
+      "epoch": 148.44,
+      "grad_norm": 14.659634590148926,
+      "learning_rate": 5.052052052052052e-06,
+      "loss": 0.3777,
+      "step": 49430
+    },
+    {
+      "epoch": 148.47,
+      "grad_norm": 11.937125205993652,
+      "learning_rate": 5.051051051051051e-06,
+      "loss": 0.4314,
+      "step": 49440
+    },
+    {
+      "epoch": 148.5,
+      "grad_norm": 19.414339065551758,
+      "learning_rate": 5.05005005005005e-06,
+      "loss": 0.4512,
+      "step": 49450
+    },
+    {
+      "epoch": 148.53,
+      "grad_norm": 14.354872703552246,
+      "learning_rate": 5.04904904904905e-06,
+      "loss": 0.3826,
+      "step": 49460
+    },
+    {
+      "epoch": 148.56,
+      "grad_norm": 14.176769256591797,
+      "learning_rate": 5.048048048048048e-06,
+      "loss": 0.3744,
+      "step": 49470
+    },
+    {
+      "epoch": 148.59,
+      "grad_norm": 11.1133394241333,
+      "learning_rate": 5.047047047047048e-06,
+      "loss": 0.3807,
+      "step": 49480
+    },
+    {
+      "epoch": 148.62,
+      "grad_norm": 13.254918098449707,
+      "learning_rate": 5.046046046046047e-06,
+      "loss": 0.4738,
+      "step": 49490
+    },
+    {
+      "epoch": 148.65,
+      "grad_norm": 16.428377151489258,
+      "learning_rate": 5.045045045045045e-06,
+      "loss": 0.3785,
+      "step": 49500
+    },
+    {
+      "epoch": 148.68,
+      "grad_norm": 19.617048263549805,
+      "learning_rate": 5.0440440440440444e-06,
+      "loss": 0.3892,
+      "step": 49510
+    },
+    {
+      "epoch": 148.71,
+      "grad_norm": 12.818624496459961,
+      "learning_rate": 5.043043043043043e-06,
+      "loss": 0.3461,
+      "step": 49520
+    },
+    {
+      "epoch": 148.74,
+      "grad_norm": 25.996417999267578,
+      "learning_rate": 5.0420420420420425e-06,
+      "loss": 0.4743,
+      "step": 49530
+    },
+    {
+      "epoch": 148.77,
+      "grad_norm": 14.286713600158691,
+      "learning_rate": 5.041041041041042e-06,
+      "loss": 0.3925,
+      "step": 49540
+    },
+    {
+      "epoch": 148.8,
+      "grad_norm": 18.605600357055664,
+      "learning_rate": 5.0400400400400405e-06,
+      "loss": 0.3966,
+      "step": 49550
+    },
+    {
+      "epoch": 148.83,
+      "grad_norm": 13.501989364624023,
+      "learning_rate": 5.03903903903904e-06,
+      "loss": 0.4068,
+      "step": 49560
+    },
+    {
+      "epoch": 148.86,
+      "grad_norm": 24.034608840942383,
+      "learning_rate": 5.0380380380380386e-06,
+      "loss": 0.3757,
+      "step": 49570
+    },
+    {
+      "epoch": 148.89,
+      "grad_norm": 17.121849060058594,
+      "learning_rate": 5.037037037037037e-06,
+      "loss": 0.4203,
+      "step": 49580
+    },
+    {
+      "epoch": 148.92,
+      "grad_norm": 16.566509246826172,
+      "learning_rate": 5.036036036036036e-06,
+      "loss": 0.3497,
+      "step": 49590
+    },
+    {
+      "epoch": 148.95,
+      "grad_norm": 11.099105834960938,
+      "learning_rate": 5.035035035035035e-06,
+      "loss": 0.3866,
+      "step": 49600
+    },
+    {
+      "epoch": 148.98,
+      "grad_norm": 13.719161033630371,
+      "learning_rate": 5.034034034034035e-06,
+      "loss": 0.3833,
+      "step": 49610
+    },
+    {
+      "epoch": 149.0,
+      "eval_accuracy": 0.865,
+      "eval_loss": 0.5228223204612732,
+      "eval_runtime": 12.8303,
+      "eval_samples_per_second": 779.402,
+      "eval_steps_per_second": 3.118,
+      "step": 49617
+    },
+    {
+      "epoch": 149.01,
+      "grad_norm": 15.503888130187988,
+      "learning_rate": 5.033033033033033e-06,
+      "loss": 0.4428,
+      "step": 49620
+    },
+    {
+      "epoch": 149.04,
+      "grad_norm": 16.798320770263672,
+      "learning_rate": 5.032032032032033e-06,
+      "loss": 0.4274,
+      "step": 49630
+    },
+    {
+      "epoch": 149.07,
+      "grad_norm": 18.786827087402344,
+      "learning_rate": 5.031031031031032e-06,
+      "loss": 0.3608,
+      "step": 49640
+    },
+    {
+      "epoch": 149.1,
+      "grad_norm": 13.984582901000977,
+      "learning_rate": 5.030030030030031e-06,
+      "loss": 0.3852,
+      "step": 49650
+    },
+    {
+      "epoch": 149.13,
+      "grad_norm": 11.216075897216797,
+      "learning_rate": 5.029029029029029e-06,
+      "loss": 0.3803,
+      "step": 49660
+    },
+    {
+      "epoch": 149.16,
+      "grad_norm": 17.6729679107666,
+      "learning_rate": 5.028028028028028e-06,
+      "loss": 0.3373,
+      "step": 49670
+    },
+    {
+      "epoch": 149.19,
+      "grad_norm": 11.334756851196289,
+      "learning_rate": 5.027027027027027e-06,
+      "loss": 0.3832,
+      "step": 49680
+    },
+    {
+      "epoch": 149.22,
+      "grad_norm": 11.469735145568848,
+      "learning_rate": 5.026026026026026e-06,
+      "loss": 0.4401,
+      "step": 49690
+    },
+    {
+      "epoch": 149.25,
+      "grad_norm": 21.279067993164062,
+      "learning_rate": 5.025025025025025e-06,
+      "loss": 0.3809,
+      "step": 49700
+    },
+    {
+      "epoch": 149.28,
+      "grad_norm": 17.329675674438477,
+      "learning_rate": 5.024024024024025e-06,
+      "loss": 0.4079,
+      "step": 49710
+    },
+    {
+      "epoch": 149.31,
+      "grad_norm": 23.64842987060547,
+      "learning_rate": 5.023023023023023e-06,
+      "loss": 0.4143,
+      "step": 49720
+    },
+    {
+      "epoch": 149.34,
+      "grad_norm": 21.753952026367188,
+      "learning_rate": 5.022022022022023e-06,
+      "loss": 0.3839,
+      "step": 49730
+    },
+    {
+      "epoch": 149.37,
+      "grad_norm": 13.854531288146973,
+      "learning_rate": 5.021021021021022e-06,
+      "loss": 0.356,
+      "step": 49740
+    },
+    {
+      "epoch": 149.4,
+      "grad_norm": 15.362092018127441,
+      "learning_rate": 5.02002002002002e-06,
+      "loss": 0.3919,
+      "step": 49750
+    },
+    {
+      "epoch": 149.43,
+      "grad_norm": 11.715888023376465,
+      "learning_rate": 5.0190190190190195e-06,
+      "loss": 0.3752,
+      "step": 49760
+    },
+    {
+      "epoch": 149.46,
+      "grad_norm": 15.008028984069824,
+      "learning_rate": 5.018018018018018e-06,
+      "loss": 0.362,
+      "step": 49770
+    },
+    {
+      "epoch": 149.49,
+      "grad_norm": 17.18223762512207,
+      "learning_rate": 5.0170170170170175e-06,
+      "loss": 0.4057,
+      "step": 49780
+    },
+    {
+      "epoch": 149.52,
+      "grad_norm": 14.067462921142578,
+      "learning_rate": 5.016016016016016e-06,
+      "loss": 0.4019,
+      "step": 49790
+    },
+    {
+      "epoch": 149.55,
+      "grad_norm": 14.481401443481445,
+      "learning_rate": 5.0150150150150156e-06,
+      "loss": 0.3803,
+      "step": 49800
+    },
+    {
+      "epoch": 149.58,
+      "grad_norm": 18.524555206298828,
+      "learning_rate": 5.014014014014015e-06,
+      "loss": 0.4595,
+      "step": 49810
+    },
+    {
+      "epoch": 149.61,
+      "grad_norm": 17.826330184936523,
+      "learning_rate": 5.013013013013014e-06,
+      "loss": 0.4001,
+      "step": 49820
+    },
+    {
+      "epoch": 149.64,
+      "grad_norm": 13.367208480834961,
+      "learning_rate": 5.012012012012012e-06,
+      "loss": 0.3541,
+      "step": 49830
+    },
+    {
+      "epoch": 149.67,
+      "grad_norm": 13.676459312438965,
+      "learning_rate": 5.011011011011011e-06,
+      "loss": 0.3833,
+      "step": 49840
+    },
+    {
+      "epoch": 149.7,
+      "grad_norm": 18.194246292114258,
+      "learning_rate": 5.01001001001001e-06,
+      "loss": 0.4266,
+      "step": 49850
+    },
+    {
+      "epoch": 149.73,
+      "grad_norm": 22.165494918823242,
+      "learning_rate": 5.00900900900901e-06,
+      "loss": 0.3831,
+      "step": 49860
+    },
+    {
+      "epoch": 149.76,
+      "grad_norm": 17.494686126708984,
+      "learning_rate": 5.008008008008008e-06,
+      "loss": 0.4004,
+      "step": 49870
+    },
+    {
+      "epoch": 149.79,
+      "grad_norm": 14.398700714111328,
+      "learning_rate": 5.007007007007008e-06,
+      "loss": 0.3717,
+      "step": 49880
+    },
+    {
+      "epoch": 149.82,
+      "grad_norm": 16.39318084716797,
+      "learning_rate": 5.006006006006007e-06,
+      "loss": 0.4121,
+      "step": 49890
+    },
+    {
+      "epoch": 149.85,
+      "grad_norm": 16.817285537719727,
+      "learning_rate": 5.005005005005006e-06,
+      "loss": 0.3856,
+      "step": 49900
+    },
+    {
+      "epoch": 149.88,
+      "grad_norm": 21.776582717895508,
+      "learning_rate": 5.0040040040040035e-06,
+      "loss": 0.3681,
+      "step": 49910
+    },
+    {
+      "epoch": 149.91,
+      "grad_norm": 12.703767776489258,
+      "learning_rate": 5.003003003003003e-06,
+      "loss": 0.4124,
+      "step": 49920
+    },
+    {
+      "epoch": 149.94,
+      "grad_norm": 27.493621826171875,
+      "learning_rate": 5.002002002002002e-06,
+      "loss": 0.3829,
+      "step": 49930
+    },
+    {
+      "epoch": 149.97,
+      "grad_norm": 12.713025093078613,
+      "learning_rate": 5.001001001001001e-06,
+      "loss": 0.3866,
+      "step": 49940
+    },
+    {
+      "epoch": 150.0,
+      "grad_norm": 0.878166675567627,
+      "learning_rate": 5e-06,
+      "loss": 0.3934,
+      "step": 49950
+    },
+    {
+      "epoch": 150.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5222606658935547,
+      "eval_runtime": 12.8233,
+      "eval_samples_per_second": 779.829,
+      "eval_steps_per_second": 3.119,
+      "step": 49950
+    },
+    {
+      "epoch": 150.03,
+      "grad_norm": 12.509446144104004,
+      "learning_rate": 4.998998998999e-06,
+      "loss": 0.3703,
+      "step": 49960
+    },
+    {
+      "epoch": 150.06,
+      "grad_norm": 28.933643341064453,
+      "learning_rate": 4.9979979979979984e-06,
+      "loss": 0.388,
+      "step": 49970
+    },
+    {
+      "epoch": 150.09,
+      "grad_norm": 12.608945846557617,
+      "learning_rate": 4.996996996996997e-06,
+      "loss": 0.4307,
+      "step": 49980
+    },
+    {
+      "epoch": 150.12,
+      "grad_norm": 10.060328483581543,
+      "learning_rate": 4.9959959959959965e-06,
+      "loss": 0.3955,
+      "step": 49990
+    },
+    {
+      "epoch": 150.15,
+      "grad_norm": 13.228853225708008,
+      "learning_rate": 4.994994994994996e-06,
+      "loss": 0.3794,
+      "step": 50000
+    },
+    {
+      "epoch": 150.18,
+      "grad_norm": 13.416147232055664,
+      "learning_rate": 4.9939939939939945e-06,
+      "loss": 0.4255,
+      "step": 50010
+    },
+    {
+      "epoch": 150.21,
+      "grad_norm": 20.14986801147461,
+      "learning_rate": 4.992992992992993e-06,
+      "loss": 0.4193,
+      "step": 50020
+    },
+    {
+      "epoch": 150.24,
+      "grad_norm": 21.813894271850586,
+      "learning_rate": 4.9919919919919926e-06,
+      "loss": 0.3874,
+      "step": 50030
+    },
+    {
+      "epoch": 150.27,
+      "grad_norm": 18.33965301513672,
+      "learning_rate": 4.990990990990991e-06,
+      "loss": 0.3908,
+      "step": 50040
+    },
+    {
+      "epoch": 150.3,
+      "grad_norm": 21.07203483581543,
+      "learning_rate": 4.989989989989991e-06,
+      "loss": 0.4181,
+      "step": 50050
+    },
+    {
+      "epoch": 150.33,
+      "grad_norm": 11.073260307312012,
+      "learning_rate": 4.988988988988989e-06,
+      "loss": 0.4225,
+      "step": 50060
+    },
+    {
+      "epoch": 150.36,
+      "grad_norm": 14.550000190734863,
+      "learning_rate": 4.987987987987989e-06,
+      "loss": 0.4163,
+      "step": 50070
+    },
+    {
+      "epoch": 150.39,
+      "grad_norm": 15.44798469543457,
+      "learning_rate": 4.986986986986987e-06,
+      "loss": 0.3767,
+      "step": 50080
+    },
+    {
+      "epoch": 150.42,
+      "grad_norm": 13.305105209350586,
+      "learning_rate": 4.985985985985987e-06,
+      "loss": 0.3501,
+      "step": 50090
+    },
+    {
+      "epoch": 150.45,
+      "grad_norm": 18.041728973388672,
+      "learning_rate": 4.984984984984985e-06,
+      "loss": 0.4032,
+      "step": 50100
+    },
+    {
+      "epoch": 150.48,
+      "grad_norm": 19.850709915161133,
+      "learning_rate": 4.983983983983985e-06,
+      "loss": 0.3997,
+      "step": 50110
+    },
+    {
+      "epoch": 150.51,
+      "grad_norm": 21.864816665649414,
+      "learning_rate": 4.982982982982983e-06,
+      "loss": 0.3834,
+      "step": 50120
+    },
+    {
+      "epoch": 150.54,
+      "grad_norm": 15.840153694152832,
+      "learning_rate": 4.981981981981983e-06,
+      "loss": 0.3843,
+      "step": 50130
+    },
+    {
+      "epoch": 150.57,
+      "grad_norm": 17.679149627685547,
+      "learning_rate": 4.980980980980981e-06,
+      "loss": 0.423,
+      "step": 50140
+    },
+    {
+      "epoch": 150.6,
+      "grad_norm": 13.276403427124023,
+      "learning_rate": 4.97997997997998e-06,
+      "loss": 0.4161,
+      "step": 50150
+    },
+    {
+      "epoch": 150.63,
+      "grad_norm": 19.15127944946289,
+      "learning_rate": 4.978978978978979e-06,
+      "loss": 0.4201,
+      "step": 50160
+    },
+    {
+      "epoch": 150.66,
+      "grad_norm": 17.688386917114258,
+      "learning_rate": 4.977977977977979e-06,
+      "loss": 0.4079,
+      "step": 50170
+    },
+    {
+      "epoch": 150.69,
+      "grad_norm": 13.520710945129395,
+      "learning_rate": 4.976976976976977e-06,
+      "loss": 0.3549,
+      "step": 50180
+    },
+    {
+      "epoch": 150.72,
+      "grad_norm": 16.257078170776367,
+      "learning_rate": 4.975975975975976e-06,
+      "loss": 0.4199,
+      "step": 50190
+    },
+    {
+      "epoch": 150.75,
+      "grad_norm": 16.796720504760742,
+      "learning_rate": 4.9749749749749754e-06,
+      "loss": 0.4481,
+      "step": 50200
+    },
+    {
+      "epoch": 150.78,
+      "grad_norm": 14.490126609802246,
+      "learning_rate": 4.973973973973975e-06,
+      "loss": 0.408,
+      "step": 50210
+    },
+    {
+      "epoch": 150.81,
+      "grad_norm": 17.615772247314453,
+      "learning_rate": 4.9729729729729735e-06,
+      "loss": 0.4068,
+      "step": 50220
+    },
+    {
+      "epoch": 150.84,
+      "grad_norm": 12.751334190368652,
+      "learning_rate": 4.971971971971972e-06,
+      "loss": 0.3992,
+      "step": 50230
+    },
+    {
+      "epoch": 150.87,
+      "grad_norm": 21.668352127075195,
+      "learning_rate": 4.9709709709709715e-06,
+      "loss": 0.4313,
+      "step": 50240
+    },
+    {
+      "epoch": 150.9,
+      "grad_norm": 20.70770263671875,
+      "learning_rate": 4.96996996996997e-06,
+      "loss": 0.4477,
+      "step": 50250
+    },
+    {
+      "epoch": 150.93,
+      "grad_norm": 16.957801818847656,
+      "learning_rate": 4.9689689689689696e-06,
+      "loss": 0.3988,
+      "step": 50260
+    },
+    {
+      "epoch": 150.96,
+      "grad_norm": 14.51667594909668,
+      "learning_rate": 4.967967967967968e-06,
+      "loss": 0.3641,
+      "step": 50270
+    },
+    {
+      "epoch": 150.99,
+      "grad_norm": 25.423946380615234,
+      "learning_rate": 4.966966966966968e-06,
+      "loss": 0.4226,
+      "step": 50280
+    },
+    {
+      "epoch": 151.0,
+      "eval_accuracy": 0.8657,
+      "eval_loss": 0.5183530449867249,
+      "eval_runtime": 12.4299,
+      "eval_samples_per_second": 804.513,
+      "eval_steps_per_second": 3.218,
+      "step": 50283
+    },
+    {
+      "epoch": 151.02,
+      "grad_norm": 13.849394798278809,
+      "learning_rate": 4.965965965965966e-06,
+      "loss": 0.3989,
+      "step": 50290
+    },
+    {
+      "epoch": 151.05,
+      "grad_norm": 13.99044418334961,
+      "learning_rate": 4.964964964964966e-06,
+      "loss": 0.4121,
+      "step": 50300
+    },
+    {
+      "epoch": 151.08,
+      "grad_norm": 15.256455421447754,
+      "learning_rate": 4.963963963963964e-06,
+      "loss": 0.3696,
+      "step": 50310
+    },
+    {
+      "epoch": 151.11,
+      "grad_norm": 22.821115493774414,
+      "learning_rate": 4.962962962962964e-06,
+      "loss": 0.4104,
+      "step": 50320
+    },
+    {
+      "epoch": 151.14,
+      "grad_norm": 18.970617294311523,
+      "learning_rate": 4.961961961961962e-06,
+      "loss": 0.4025,
+      "step": 50330
+    },
+    {
+      "epoch": 151.17,
+      "grad_norm": 15.599268913269043,
+      "learning_rate": 4.960960960960962e-06,
+      "loss": 0.406,
+      "step": 50340
+    },
+    {
+      "epoch": 151.2,
+      "grad_norm": 21.71476173400879,
+      "learning_rate": 4.95995995995996e-06,
+      "loss": 0.4039,
+      "step": 50350
+    },
+    {
+      "epoch": 151.23,
+      "grad_norm": 20.88029670715332,
+      "learning_rate": 4.958958958958959e-06,
+      "loss": 0.4113,
+      "step": 50360
+    },
+    {
+      "epoch": 151.26,
+      "grad_norm": 14.431530952453613,
+      "learning_rate": 4.957957957957958e-06,
+      "loss": 0.3914,
+      "step": 50370
+    },
+    {
+      "epoch": 151.29,
+      "grad_norm": 13.027268409729004,
+      "learning_rate": 4.956956956956958e-06,
+      "loss": 0.3951,
+      "step": 50380
+    },
+    {
+      "epoch": 151.32,
+      "grad_norm": 9.745125770568848,
+      "learning_rate": 4.955955955955956e-06,
+      "loss": 0.3318,
+      "step": 50390
+    },
+    {
+      "epoch": 151.35,
+      "grad_norm": 11.522449493408203,
+      "learning_rate": 4.954954954954955e-06,
+      "loss": 0.3514,
+      "step": 50400
+    },
+    {
+      "epoch": 151.38,
+      "grad_norm": 17.34370231628418,
+      "learning_rate": 4.953953953953954e-06,
+      "loss": 0.3807,
+      "step": 50410
+    },
+    {
+      "epoch": 151.41,
+      "grad_norm": 29.007287979125977,
+      "learning_rate": 4.952952952952954e-06,
+      "loss": 0.4375,
+      "step": 50420
+    },
+    {
+      "epoch": 151.44,
+      "grad_norm": 16.62506866455078,
+      "learning_rate": 4.9519519519519524e-06,
+      "loss": 0.4177,
+      "step": 50430
+    },
+    {
+      "epoch": 151.47,
+      "grad_norm": 29.575214385986328,
+      "learning_rate": 4.950950950950951e-06,
+      "loss": 0.3669,
+      "step": 50440
+    },
+    {
+      "epoch": 151.5,
+      "grad_norm": 14.662795066833496,
+      "learning_rate": 4.9499499499499505e-06,
+      "loss": 0.3813,
+      "step": 50450
+    },
+    {
+      "epoch": 151.53,
+      "grad_norm": 12.983796119689941,
+      "learning_rate": 4.94894894894895e-06,
+      "loss": 0.3873,
+      "step": 50460
+    },
+    {
+      "epoch": 151.56,
+      "grad_norm": 13.419620513916016,
+      "learning_rate": 4.947947947947948e-06,
+      "loss": 0.4275,
+      "step": 50470
+    },
+    {
+      "epoch": 151.59,
+      "grad_norm": 20.96712875366211,
+      "learning_rate": 4.946946946946947e-06,
+      "loss": 0.4366,
+      "step": 50480
+    },
+    {
+      "epoch": 151.62,
+      "grad_norm": 15.87871265411377,
+      "learning_rate": 4.9459459459459466e-06,
+      "loss": 0.3894,
+      "step": 50490
+    },
+    {
+      "epoch": 151.65,
+      "grad_norm": 12.118277549743652,
+      "learning_rate": 4.944944944944945e-06,
+      "loss": 0.4083,
+      "step": 50500
+    },
+    {
+      "epoch": 151.68,
+      "grad_norm": 18.364442825317383,
+      "learning_rate": 4.943943943943945e-06,
+      "loss": 0.3677,
+      "step": 50510
+    },
+    {
+      "epoch": 151.71,
+      "grad_norm": 13.885570526123047,
+      "learning_rate": 4.942942942942943e-06,
+      "loss": 0.4129,
+      "step": 50520
+    },
+    {
+      "epoch": 151.74,
+      "grad_norm": 19.197553634643555,
+      "learning_rate": 4.941941941941943e-06,
+      "loss": 0.4063,
+      "step": 50530
+    },
+    {
+      "epoch": 151.77,
+      "grad_norm": 19.43085289001465,
+      "learning_rate": 4.940940940940941e-06,
+      "loss": 0.3772,
+      "step": 50540
+    },
+    {
+      "epoch": 151.8,
+      "grad_norm": 13.23403263092041,
+      "learning_rate": 4.939939939939941e-06,
+      "loss": 0.4206,
+      "step": 50550
+    },
+    {
+      "epoch": 151.83,
+      "grad_norm": 20.52122688293457,
+      "learning_rate": 4.938938938938939e-06,
+      "loss": 0.4425,
+      "step": 50560
+    },
+    {
+      "epoch": 151.86,
+      "grad_norm": 9.390904426574707,
+      "learning_rate": 4.937937937937939e-06,
+      "loss": 0.3512,
+      "step": 50570
+    },
+    {
+      "epoch": 151.89,
+      "grad_norm": 11.063456535339355,
+      "learning_rate": 4.936936936936937e-06,
+      "loss": 0.3869,
+      "step": 50580
+    },
+    {
+      "epoch": 151.92,
+      "grad_norm": 14.061050415039062,
+      "learning_rate": 4.935935935935937e-06,
+      "loss": 0.3791,
+      "step": 50590
+    },
+    {
+      "epoch": 151.95,
+      "grad_norm": 20.17213249206543,
+      "learning_rate": 4.934934934934935e-06,
+      "loss": 0.3679,
+      "step": 50600
+    },
+    {
+      "epoch": 151.98,
+      "grad_norm": 15.663177490234375,
+      "learning_rate": 4.933933933933934e-06,
+      "loss": 0.4112,
+      "step": 50610
+    },
+    {
+      "epoch": 152.0,
+      "eval_accuracy": 0.8644,
+      "eval_loss": 0.5186694264411926,
+      "eval_runtime": 12.9009,
+      "eval_samples_per_second": 775.14,
+      "eval_steps_per_second": 3.101,
+      "step": 50616
+    },
+    {
+      "epoch": 152.01,
+      "grad_norm": 13.139857292175293,
+      "learning_rate": 4.932932932932933e-06,
+      "loss": 0.4349,
+      "step": 50620
+    },
+    {
+      "epoch": 152.04,
+      "grad_norm": 9.718353271484375,
+      "learning_rate": 4.931931931931933e-06,
+      "loss": 0.3684,
+      "step": 50630
+    },
+    {
+      "epoch": 152.07,
+      "grad_norm": 22.924530029296875,
+      "learning_rate": 4.930930930930931e-06,
+      "loss": 0.4013,
+      "step": 50640
+    },
+    {
+      "epoch": 152.1,
+      "grad_norm": 11.234991073608398,
+      "learning_rate": 4.92992992992993e-06,
+      "loss": 0.3562,
+      "step": 50650
+    },
+    {
+      "epoch": 152.13,
+      "grad_norm": 17.012784957885742,
+      "learning_rate": 4.9289289289289294e-06,
+      "loss": 0.3714,
+      "step": 50660
+    },
+    {
+      "epoch": 152.16,
+      "grad_norm": 16.58921241760254,
+      "learning_rate": 4.927927927927929e-06,
+      "loss": 0.3596,
+      "step": 50670
+    },
+    {
+      "epoch": 152.19,
+      "grad_norm": 12.964404106140137,
+      "learning_rate": 4.9269269269269275e-06,
+      "loss": 0.4113,
+      "step": 50680
+    },
+    {
+      "epoch": 152.22,
+      "grad_norm": 19.07688331604004,
+      "learning_rate": 4.925925925925926e-06,
+      "loss": 0.3802,
+      "step": 50690
+    },
+    {
+      "epoch": 152.25,
+      "grad_norm": 21.89029884338379,
+      "learning_rate": 4.9249249249249255e-06,
+      "loss": 0.3642,
+      "step": 50700
+    },
+    {
+      "epoch": 152.28,
+      "grad_norm": 17.81185531616211,
+      "learning_rate": 4.923923923923924e-06,
+      "loss": 0.4051,
+      "step": 50710
+    },
+    {
+      "epoch": 152.31,
+      "grad_norm": 16.545194625854492,
+      "learning_rate": 4.922922922922923e-06,
+      "loss": 0.4282,
+      "step": 50720
+    },
+    {
+      "epoch": 152.34,
+      "grad_norm": 12.610400199890137,
+      "learning_rate": 4.921921921921922e-06,
+      "loss": 0.4345,
+      "step": 50730
+    },
+    {
+      "epoch": 152.37,
+      "grad_norm": 18.83820343017578,
+      "learning_rate": 4.920920920920922e-06,
+      "loss": 0.3448,
+      "step": 50740
+    },
+    {
+      "epoch": 152.4,
+      "grad_norm": 12.140055656433105,
+      "learning_rate": 4.91991991991992e-06,
+      "loss": 0.3992,
+      "step": 50750
+    },
+    {
+      "epoch": 152.43,
+      "grad_norm": 18.246797561645508,
+      "learning_rate": 4.91891891891892e-06,
+      "loss": 0.4132,
+      "step": 50760
+    },
+    {
+      "epoch": 152.46,
+      "grad_norm": 17.833351135253906,
+      "learning_rate": 4.917917917917918e-06,
+      "loss": 0.4169,
+      "step": 50770
+    },
+    {
+      "epoch": 152.49,
+      "grad_norm": 15.930706977844238,
+      "learning_rate": 4.916916916916918e-06,
+      "loss": 0.3409,
+      "step": 50780
+    },
+    {
+      "epoch": 152.52,
+      "grad_norm": 12.83889102935791,
+      "learning_rate": 4.915915915915916e-06,
+      "loss": 0.3728,
+      "step": 50790
+    },
+    {
+      "epoch": 152.55,
+      "grad_norm": 21.39049530029297,
+      "learning_rate": 4.914914914914916e-06,
+      "loss": 0.4025,
+      "step": 50800
+    },
+    {
+      "epoch": 152.58,
+      "grad_norm": 23.8992862701416,
+      "learning_rate": 4.913913913913914e-06,
+      "loss": 0.3894,
+      "step": 50810
+    },
+    {
+      "epoch": 152.61,
+      "grad_norm": 13.00473403930664,
+      "learning_rate": 4.912912912912913e-06,
+      "loss": 0.4104,
+      "step": 50820
+    },
+    {
+      "epoch": 152.64,
+      "grad_norm": 11.818963050842285,
+      "learning_rate": 4.911911911911912e-06,
+      "loss": 0.3484,
+      "step": 50830
+    },
+    {
+      "epoch": 152.67,
+      "grad_norm": 13.102164268493652,
+      "learning_rate": 4.910910910910912e-06,
+      "loss": 0.4024,
+      "step": 50840
+    },
+    {
+      "epoch": 152.7,
+      "grad_norm": 10.22575855255127,
+      "learning_rate": 4.90990990990991e-06,
+      "loss": 0.3909,
+      "step": 50850
+    },
+    {
+      "epoch": 152.73,
+      "grad_norm": 14.699150085449219,
+      "learning_rate": 4.908908908908909e-06,
+      "loss": 0.4149,
+      "step": 50860
+    },
+    {
+      "epoch": 152.76,
+      "grad_norm": 33.110076904296875,
+      "learning_rate": 4.907907907907908e-06,
+      "loss": 0.4182,
+      "step": 50870
+    },
+    {
+      "epoch": 152.79,
+      "grad_norm": 15.751762390136719,
+      "learning_rate": 4.906906906906908e-06,
+      "loss": 0.3882,
+      "step": 50880
+    },
+    {
+      "epoch": 152.82,
+      "grad_norm": 16.7052059173584,
+      "learning_rate": 4.9059059059059064e-06,
+      "loss": 0.3825,
+      "step": 50890
+    },
+    {
+      "epoch": 152.85,
+      "grad_norm": 19.044137954711914,
+      "learning_rate": 4.904904904904905e-06,
+      "loss": 0.4113,
+      "step": 50900
+    },
+    {
+      "epoch": 152.88,
+      "grad_norm": 21.486915588378906,
+      "learning_rate": 4.9039039039039045e-06,
+      "loss": 0.3915,
+      "step": 50910
+    },
+    {
+      "epoch": 152.91,
+      "grad_norm": 13.558988571166992,
+      "learning_rate": 4.902902902902904e-06,
+      "loss": 0.3388,
+      "step": 50920
+    },
+    {
+      "epoch": 152.94,
+      "grad_norm": 22.816499710083008,
+      "learning_rate": 4.901901901901902e-06,
+      "loss": 0.382,
+      "step": 50930
+    },
+    {
+      "epoch": 152.97,
+      "grad_norm": 15.753442764282227,
+      "learning_rate": 4.900900900900901e-06,
+      "loss": 0.4202,
+      "step": 50940
+    },
+    {
+      "epoch": 153.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5191443562507629,
+      "eval_runtime": 12.7235,
+      "eval_samples_per_second": 785.947,
+      "eval_steps_per_second": 3.144,
+      "step": 50949
+    },
+    {
+      "epoch": 153.0,
+      "grad_norm": 18.10231590270996,
+      "learning_rate": 4.8998998998999005e-06,
+      "loss": 0.3855,
+      "step": 50950
+    },
+    {
+      "epoch": 153.03,
+      "grad_norm": 17.266984939575195,
+      "learning_rate": 4.898898898898899e-06,
+      "loss": 0.4361,
+      "step": 50960
+    },
+    {
+      "epoch": 153.06,
+      "grad_norm": 17.26080322265625,
+      "learning_rate": 4.897897897897898e-06,
+      "loss": 0.3568,
+      "step": 50970
+    },
+    {
+      "epoch": 153.09,
+      "grad_norm": 15.20398235321045,
+      "learning_rate": 4.896896896896897e-06,
+      "loss": 0.3546,
+      "step": 50980
+    },
+    {
+      "epoch": 153.12,
+      "grad_norm": 15.783230781555176,
+      "learning_rate": 4.895895895895897e-06,
+      "loss": 0.3929,
+      "step": 50990
+    },
+    {
+      "epoch": 153.15,
+      "grad_norm": 16.53842544555664,
+      "learning_rate": 4.894894894894895e-06,
+      "loss": 0.4138,
+      "step": 51000
+    },
+    {
+      "epoch": 153.18,
+      "grad_norm": 12.86841869354248,
+      "learning_rate": 4.893893893893894e-06,
+      "loss": 0.3888,
+      "step": 51010
+    },
+    {
+      "epoch": 153.21,
+      "grad_norm": 16.251529693603516,
+      "learning_rate": 4.892892892892893e-06,
+      "loss": 0.36,
+      "step": 51020
+    },
+    {
+      "epoch": 153.24,
+      "grad_norm": 15.550637245178223,
+      "learning_rate": 4.891891891891893e-06,
+      "loss": 0.3682,
+      "step": 51030
+    },
+    {
+      "epoch": 153.27,
+      "grad_norm": 17.623355865478516,
+      "learning_rate": 4.890890890890891e-06,
+      "loss": 0.3776,
+      "step": 51040
+    },
+    {
+      "epoch": 153.3,
+      "grad_norm": 12.51807689666748,
+      "learning_rate": 4.889889889889891e-06,
+      "loss": 0.4058,
+      "step": 51050
+    },
+    {
+      "epoch": 153.33,
+      "grad_norm": 12.890047073364258,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 0.3855,
+      "step": 51060
+    },
+    {
+      "epoch": 153.36,
+      "grad_norm": 28.895051956176758,
+      "learning_rate": 4.887887887887888e-06,
+      "loss": 0.382,
+      "step": 51070
+    },
+    {
+      "epoch": 153.39,
+      "grad_norm": 15.497566223144531,
+      "learning_rate": 4.886886886886887e-06,
+      "loss": 0.4171,
+      "step": 51080
+    },
+    {
+      "epoch": 153.42,
+      "grad_norm": 15.79238510131836,
+      "learning_rate": 4.885885885885887e-06,
+      "loss": 0.4052,
+      "step": 51090
+    },
+    {
+      "epoch": 153.45,
+      "grad_norm": 12.09903621673584,
+      "learning_rate": 4.884884884884885e-06,
+      "loss": 0.4222,
+      "step": 51100
+    },
+    {
+      "epoch": 153.48,
+      "grad_norm": 15.322608947753906,
+      "learning_rate": 4.883883883883884e-06,
+      "loss": 0.4097,
+      "step": 51110
+    },
+    {
+      "epoch": 153.51,
+      "grad_norm": 15.61450481414795,
+      "learning_rate": 4.8828828828828834e-06,
+      "loss": 0.4297,
+      "step": 51120
+    },
+    {
+      "epoch": 153.54,
+      "grad_norm": 11.877705574035645,
+      "learning_rate": 4.881881881881883e-06,
+      "loss": 0.4404,
+      "step": 51130
+    },
+    {
+      "epoch": 153.57,
+      "grad_norm": 18.822420120239258,
+      "learning_rate": 4.8808808808808815e-06,
+      "loss": 0.3648,
+      "step": 51140
+    },
+    {
+      "epoch": 153.6,
+      "grad_norm": 12.988822937011719,
+      "learning_rate": 4.87987987987988e-06,
+      "loss": 0.3699,
+      "step": 51150
+    },
+    {
+      "epoch": 153.63,
+      "grad_norm": 13.715423583984375,
+      "learning_rate": 4.8788788788788795e-06,
+      "loss": 0.3655,
+      "step": 51160
+    },
+    {
+      "epoch": 153.66,
+      "grad_norm": 16.237682342529297,
+      "learning_rate": 4.877877877877878e-06,
+      "loss": 0.4246,
+      "step": 51170
+    },
+    {
+      "epoch": 153.69,
+      "grad_norm": 11.37242317199707,
+      "learning_rate": 4.876876876876877e-06,
+      "loss": 0.3878,
+      "step": 51180
+    },
+    {
+      "epoch": 153.72,
+      "grad_norm": 14.910950660705566,
+      "learning_rate": 4.875875875875876e-06,
+      "loss": 0.408,
+      "step": 51190
+    },
+    {
+      "epoch": 153.75,
+      "grad_norm": 12.674115180969238,
+      "learning_rate": 4.874874874874876e-06,
+      "loss": 0.3623,
+      "step": 51200
+    },
+    {
+      "epoch": 153.78,
+      "grad_norm": 26.625303268432617,
+      "learning_rate": 4.873873873873874e-06,
+      "loss": 0.3934,
+      "step": 51210
+    },
+    {
+      "epoch": 153.81,
+      "grad_norm": 11.09316635131836,
+      "learning_rate": 4.872872872872873e-06,
+      "loss": 0.3539,
+      "step": 51220
+    },
+    {
+      "epoch": 153.84,
+      "grad_norm": 15.366019248962402,
+      "learning_rate": 4.871871871871872e-06,
+      "loss": 0.3695,
+      "step": 51230
+    },
+    {
+      "epoch": 153.87,
+      "grad_norm": 10.745084762573242,
+      "learning_rate": 4.870870870870872e-06,
+      "loss": 0.4003,
+      "step": 51240
+    },
+    {
+      "epoch": 153.9,
+      "grad_norm": 13.67747974395752,
+      "learning_rate": 4.86986986986987e-06,
+      "loss": 0.3758,
+      "step": 51250
+    },
+    {
+      "epoch": 153.93,
+      "grad_norm": 23.01776885986328,
+      "learning_rate": 4.868868868868869e-06,
+      "loss": 0.3645,
+      "step": 51260
+    },
+    {
+      "epoch": 153.96,
+      "grad_norm": 13.4940767288208,
+      "learning_rate": 4.867867867867868e-06,
+      "loss": 0.386,
+      "step": 51270
+    },
+    {
+      "epoch": 153.99,
+      "grad_norm": 12.809427261352539,
+      "learning_rate": 4.866866866866867e-06,
+      "loss": 0.4026,
+      "step": 51280
+    },
+    {
+      "epoch": 154.0,
+      "eval_accuracy": 0.8669,
+      "eval_loss": 0.5164682865142822,
+      "eval_runtime": 12.5061,
+      "eval_samples_per_second": 799.61,
+      "eval_steps_per_second": 3.198,
+      "step": 51282
+    },
+    {
+      "epoch": 154.02,
+      "grad_norm": 14.48398208618164,
+      "learning_rate": 4.865865865865866e-06,
+      "loss": 0.3915,
+      "step": 51290
+    },
+    {
+      "epoch": 154.05,
+      "grad_norm": 14.675040245056152,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.3636,
+      "step": 51300
+    },
+    {
+      "epoch": 154.08,
+      "grad_norm": 14.634662628173828,
+      "learning_rate": 4.863863863863864e-06,
+      "loss": 0.424,
+      "step": 51310
+    },
+    {
+      "epoch": 154.11,
+      "grad_norm": 13.915486335754395,
+      "learning_rate": 4.862862862862863e-06,
+      "loss": 0.3689,
+      "step": 51320
+    },
+    {
+      "epoch": 154.14,
+      "grad_norm": 8.86575698852539,
+      "learning_rate": 4.861861861861862e-06,
+      "loss": 0.3587,
+      "step": 51330
+    },
+    {
+      "epoch": 154.17,
+      "grad_norm": 10.953817367553711,
+      "learning_rate": 4.860860860860862e-06,
+      "loss": 0.396,
+      "step": 51340
+    },
+    {
+      "epoch": 154.2,
+      "grad_norm": 19.350893020629883,
+      "learning_rate": 4.8598598598598604e-06,
+      "loss": 0.4504,
+      "step": 51350
+    },
+    {
+      "epoch": 154.23,
+      "grad_norm": 14.872597694396973,
+      "learning_rate": 4.858858858858859e-06,
+      "loss": 0.3741,
+      "step": 51360
+    },
+    {
+      "epoch": 154.26,
+      "grad_norm": 14.399934768676758,
+      "learning_rate": 4.8578578578578585e-06,
+      "loss": 0.4061,
+      "step": 51370
+    },
+    {
+      "epoch": 154.29,
+      "grad_norm": 20.154170989990234,
+      "learning_rate": 4.856856856856857e-06,
+      "loss": 0.3852,
+      "step": 51380
+    },
+    {
+      "epoch": 154.32,
+      "grad_norm": 22.10881233215332,
+      "learning_rate": 4.855855855855856e-06,
+      "loss": 0.4314,
+      "step": 51390
+    },
+    {
+      "epoch": 154.35,
+      "grad_norm": 23.983245849609375,
+      "learning_rate": 4.854854854854855e-06,
+      "loss": 0.4074,
+      "step": 51400
+    },
+    {
+      "epoch": 154.38,
+      "grad_norm": 15.803277015686035,
+      "learning_rate": 4.8538538538538545e-06,
+      "loss": 0.397,
+      "step": 51410
+    },
+    {
+      "epoch": 154.41,
+      "grad_norm": 14.65795612335205,
+      "learning_rate": 4.852852852852853e-06,
+      "loss": 0.3759,
+      "step": 51420
+    },
+    {
+      "epoch": 154.44,
+      "grad_norm": 12.135204315185547,
+      "learning_rate": 4.851851851851852e-06,
+      "loss": 0.416,
+      "step": 51430
+    },
+    {
+      "epoch": 154.47,
+      "grad_norm": 12.379916191101074,
+      "learning_rate": 4.850850850850851e-06,
+      "loss": 0.3599,
+      "step": 51440
+    },
+    {
+      "epoch": 154.5,
+      "grad_norm": 17.605741500854492,
+      "learning_rate": 4.849849849849851e-06,
+      "loss": 0.3272,
+      "step": 51450
+    },
+    {
+      "epoch": 154.53,
+      "grad_norm": 13.077641487121582,
+      "learning_rate": 4.848848848848849e-06,
+      "loss": 0.4011,
+      "step": 51460
+    },
+    {
+      "epoch": 154.56,
+      "grad_norm": 13.617865562438965,
+      "learning_rate": 4.847847847847848e-06,
+      "loss": 0.4459,
+      "step": 51470
+    },
+    {
+      "epoch": 154.59,
+      "grad_norm": 17.416555404663086,
+      "learning_rate": 4.846846846846847e-06,
+      "loss": 0.417,
+      "step": 51480
+    },
+    {
+      "epoch": 154.62,
+      "grad_norm": 18.33196258544922,
+      "learning_rate": 4.845845845845847e-06,
+      "loss": 0.4402,
+      "step": 51490
+    },
+    {
+      "epoch": 154.65,
+      "grad_norm": 10.2378511428833,
+      "learning_rate": 4.844844844844845e-06,
+      "loss": 0.4344,
+      "step": 51500
+    },
+    {
+      "epoch": 154.68,
+      "grad_norm": 31.183429718017578,
+      "learning_rate": 4.843843843843844e-06,
+      "loss": 0.3899,
+      "step": 51510
+    },
+    {
+      "epoch": 154.71,
+      "grad_norm": 13.957683563232422,
+      "learning_rate": 4.842842842842843e-06,
+      "loss": 0.3791,
+      "step": 51520
+    },
+    {
+      "epoch": 154.74,
+      "grad_norm": 15.05156421661377,
+      "learning_rate": 4.841841841841842e-06,
+      "loss": 0.4142,
+      "step": 51530
+    },
+    {
+      "epoch": 154.77,
+      "grad_norm": 12.831008911132812,
+      "learning_rate": 4.840840840840841e-06,
+      "loss": 0.3653,
+      "step": 51540
+    },
+    {
+      "epoch": 154.8,
+      "grad_norm": 13.544841766357422,
+      "learning_rate": 4.83983983983984e-06,
+      "loss": 0.426,
+      "step": 51550
+    },
+    {
+      "epoch": 154.83,
+      "grad_norm": 17.78261947631836,
+      "learning_rate": 4.838838838838839e-06,
+      "loss": 0.355,
+      "step": 51560
+    },
+    {
+      "epoch": 154.86,
+      "grad_norm": 16.725177764892578,
+      "learning_rate": 4.837837837837838e-06,
+      "loss": 0.3703,
+      "step": 51570
+    },
+    {
+      "epoch": 154.89,
+      "grad_norm": 10.112130165100098,
+      "learning_rate": 4.8368368368368374e-06,
+      "loss": 0.3886,
+      "step": 51580
+    },
+    {
+      "epoch": 154.92,
+      "grad_norm": 18.49744987487793,
+      "learning_rate": 4.835835835835837e-06,
+      "loss": 0.3755,
+      "step": 51590
+    },
+    {
+      "epoch": 154.95,
+      "grad_norm": 13.642705917358398,
+      "learning_rate": 4.8348348348348355e-06,
+      "loss": 0.4072,
+      "step": 51600
+    },
+    {
+      "epoch": 154.98,
+      "grad_norm": 14.328898429870605,
+      "learning_rate": 4.833833833833834e-06,
+      "loss": 0.4322,
+      "step": 51610
+    },
+    {
+      "epoch": 155.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5215239524841309,
+      "eval_runtime": 12.3418,
+      "eval_samples_per_second": 810.255,
+      "eval_steps_per_second": 3.241,
+      "step": 51615
+    },
+    {
+      "epoch": 155.02,
+      "grad_norm": 15.237321853637695,
+      "learning_rate": 4.8328328328328335e-06,
+      "loss": 0.4583,
+      "step": 51620
+    },
+    {
+      "epoch": 155.05,
+      "grad_norm": 20.406639099121094,
+      "learning_rate": 4.831831831831832e-06,
+      "loss": 0.4432,
+      "step": 51630
+    },
+    {
+      "epoch": 155.08,
+      "grad_norm": 15.58051872253418,
+      "learning_rate": 4.830830830830831e-06,
+      "loss": 0.3513,
+      "step": 51640
+    },
+    {
+      "epoch": 155.11,
+      "grad_norm": 11.413922309875488,
+      "learning_rate": 4.82982982982983e-06,
+      "loss": 0.3899,
+      "step": 51650
+    },
+    {
+      "epoch": 155.14,
+      "grad_norm": 15.86987590789795,
+      "learning_rate": 4.82882882882883e-06,
+      "loss": 0.4161,
+      "step": 51660
+    },
+    {
+      "epoch": 155.17,
+      "grad_norm": 14.049835205078125,
+      "learning_rate": 4.827827827827828e-06,
+      "loss": 0.3688,
+      "step": 51670
+    },
+    {
+      "epoch": 155.2,
+      "grad_norm": 18.415346145629883,
+      "learning_rate": 4.826826826826827e-06,
+      "loss": 0.4483,
+      "step": 51680
+    },
+    {
+      "epoch": 155.23,
+      "grad_norm": 13.496849060058594,
+      "learning_rate": 4.825825825825826e-06,
+      "loss": 0.3649,
+      "step": 51690
+    },
+    {
+      "epoch": 155.26,
+      "grad_norm": 11.115533828735352,
+      "learning_rate": 4.824824824824826e-06,
+      "loss": 0.3529,
+      "step": 51700
+    },
+    {
+      "epoch": 155.29,
+      "grad_norm": 13.354290962219238,
+      "learning_rate": 4.823823823823824e-06,
+      "loss": 0.3747,
+      "step": 51710
+    },
+    {
+      "epoch": 155.32,
+      "grad_norm": 17.1982364654541,
+      "learning_rate": 4.822822822822823e-06,
+      "loss": 0.407,
+      "step": 51720
+    },
+    {
+      "epoch": 155.35,
+      "grad_norm": 23.344411849975586,
+      "learning_rate": 4.821821821821822e-06,
+      "loss": 0.3694,
+      "step": 51730
+    },
+    {
+      "epoch": 155.38,
+      "grad_norm": 16.07171630859375,
+      "learning_rate": 4.820820820820821e-06,
+      "loss": 0.4243,
+      "step": 51740
+    },
+    {
+      "epoch": 155.41,
+      "grad_norm": 17.10759925842285,
+      "learning_rate": 4.81981981981982e-06,
+      "loss": 0.3892,
+      "step": 51750
+    },
+    {
+      "epoch": 155.44,
+      "grad_norm": 17.824745178222656,
+      "learning_rate": 4.818818818818819e-06,
+      "loss": 0.39,
+      "step": 51760
+    },
+    {
+      "epoch": 155.47,
+      "grad_norm": 10.845376968383789,
+      "learning_rate": 4.817817817817818e-06,
+      "loss": 0.3839,
+      "step": 51770
+    },
+    {
+      "epoch": 155.5,
+      "grad_norm": 15.731200218200684,
+      "learning_rate": 4.816816816816817e-06,
+      "loss": 0.4201,
+      "step": 51780
+    },
+    {
+      "epoch": 155.53,
+      "grad_norm": 19.03276824951172,
+      "learning_rate": 4.815815815815816e-06,
+      "loss": 0.4174,
+      "step": 51790
+    },
+    {
+      "epoch": 155.56,
+      "grad_norm": 20.995777130126953,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.4004,
+      "step": 51800
+    },
+    {
+      "epoch": 155.59,
+      "grad_norm": 18.479719161987305,
+      "learning_rate": 4.8138138138138144e-06,
+      "loss": 0.384,
+      "step": 51810
+    },
+    {
+      "epoch": 155.62,
+      "grad_norm": 11.492267608642578,
+      "learning_rate": 4.812812812812813e-06,
+      "loss": 0.3477,
+      "step": 51820
+    },
+    {
+      "epoch": 155.65,
+      "grad_norm": 16.081321716308594,
+      "learning_rate": 4.8118118118118125e-06,
+      "loss": 0.3687,
+      "step": 51830
+    },
+    {
+      "epoch": 155.68,
+      "grad_norm": 16.37592315673828,
+      "learning_rate": 4.810810810810811e-06,
+      "loss": 0.3908,
+      "step": 51840
+    },
+    {
+      "epoch": 155.71,
+      "grad_norm": 19.71565818786621,
+      "learning_rate": 4.80980980980981e-06,
+      "loss": 0.3883,
+      "step": 51850
+    },
+    {
+      "epoch": 155.74,
+      "grad_norm": 18.769245147705078,
+      "learning_rate": 4.808808808808809e-06,
+      "loss": 0.3887,
+      "step": 51860
+    },
+    {
+      "epoch": 155.77,
+      "grad_norm": 17.13928985595703,
+      "learning_rate": 4.8078078078078085e-06,
+      "loss": 0.3857,
+      "step": 51870
+    },
+    {
+      "epoch": 155.8,
+      "grad_norm": 20.682064056396484,
+      "learning_rate": 4.806806806806807e-06,
+      "loss": 0.4354,
+      "step": 51880
+    },
+    {
+      "epoch": 155.83,
+      "grad_norm": 14.045702934265137,
+      "learning_rate": 4.805805805805806e-06,
+      "loss": 0.3359,
+      "step": 51890
+    },
+    {
+      "epoch": 155.86,
+      "grad_norm": 17.652976989746094,
+      "learning_rate": 4.804804804804805e-06,
+      "loss": 0.3776,
+      "step": 51900
+    },
+    {
+      "epoch": 155.89,
+      "grad_norm": 12.32742691040039,
+      "learning_rate": 4.803803803803805e-06,
+      "loss": 0.3726,
+      "step": 51910
+    },
+    {
+      "epoch": 155.92,
+      "grad_norm": 17.662574768066406,
+      "learning_rate": 4.802802802802803e-06,
+      "loss": 0.4191,
+      "step": 51920
+    },
+    {
+      "epoch": 155.95,
+      "grad_norm": 12.509106636047363,
+      "learning_rate": 4.801801801801802e-06,
+      "loss": 0.3819,
+      "step": 51930
+    },
+    {
+      "epoch": 155.98,
+      "grad_norm": 11.108885765075684,
+      "learning_rate": 4.800800800800801e-06,
+      "loss": 0.3763,
+      "step": 51940
+    },
+    {
+      "epoch": 156.0,
+      "eval_accuracy": 0.8659,
+      "eval_loss": 0.5235413908958435,
+      "eval_runtime": 12.8686,
+      "eval_samples_per_second": 777.086,
+      "eval_steps_per_second": 3.108,
+      "step": 51948
+    },
+    {
+      "epoch": 156.01,
+      "grad_norm": 17.47244644165039,
+      "learning_rate": 4.7997997997998e-06,
+      "loss": 0.3411,
+      "step": 51950
+    },
+    {
+      "epoch": 156.04,
+      "grad_norm": 14.3109769821167,
+      "learning_rate": 4.798798798798799e-06,
+      "loss": 0.3321,
+      "step": 51960
+    },
+    {
+      "epoch": 156.07,
+      "grad_norm": 22.856874465942383,
+      "learning_rate": 4.797797797797798e-06,
+      "loss": 0.371,
+      "step": 51970
+    },
+    {
+      "epoch": 156.1,
+      "grad_norm": 21.807952880859375,
+      "learning_rate": 4.796796796796797e-06,
+      "loss": 0.3498,
+      "step": 51980
+    },
+    {
+      "epoch": 156.13,
+      "grad_norm": 21.831045150756836,
+      "learning_rate": 4.795795795795796e-06,
+      "loss": 0.4249,
+      "step": 51990
+    },
+    {
+      "epoch": 156.16,
+      "grad_norm": 10.847412109375,
+      "learning_rate": 4.794794794794795e-06,
+      "loss": 0.3894,
+      "step": 52000
+    },
+    {
+      "epoch": 156.19,
+      "grad_norm": 17.47970199584961,
+      "learning_rate": 4.793793793793794e-06,
+      "loss": 0.3639,
+      "step": 52010
+    },
+    {
+      "epoch": 156.22,
+      "grad_norm": 19.176546096801758,
+      "learning_rate": 4.792792792792793e-06,
+      "loss": 0.3702,
+      "step": 52020
+    },
+    {
+      "epoch": 156.25,
+      "grad_norm": 27.749197006225586,
+      "learning_rate": 4.791791791791792e-06,
+      "loss": 0.4028,
+      "step": 52030
+    },
+    {
+      "epoch": 156.28,
+      "grad_norm": 10.91423511505127,
+      "learning_rate": 4.7907907907907914e-06,
+      "loss": 0.4372,
+      "step": 52040
+    },
+    {
+      "epoch": 156.31,
+      "grad_norm": 18.170499801635742,
+      "learning_rate": 4.78978978978979e-06,
+      "loss": 0.3949,
+      "step": 52050
+    },
+    {
+      "epoch": 156.34,
+      "grad_norm": 11.265213012695312,
+      "learning_rate": 4.7887887887887895e-06,
+      "loss": 0.4211,
+      "step": 52060
+    },
+    {
+      "epoch": 156.37,
+      "grad_norm": 12.312078475952148,
+      "learning_rate": 4.787787787787788e-06,
+      "loss": 0.4177,
+      "step": 52070
+    },
+    {
+      "epoch": 156.4,
+      "grad_norm": 12.107789039611816,
+      "learning_rate": 4.7867867867867875e-06,
+      "loss": 0.4218,
+      "step": 52080
+    },
+    {
+      "epoch": 156.43,
+      "grad_norm": 26.669456481933594,
+      "learning_rate": 4.785785785785786e-06,
+      "loss": 0.3909,
+      "step": 52090
+    },
+    {
+      "epoch": 156.46,
+      "grad_norm": 18.997394561767578,
+      "learning_rate": 4.784784784784785e-06,
+      "loss": 0.401,
+      "step": 52100
+    },
+    {
+      "epoch": 156.49,
+      "grad_norm": 11.586363792419434,
+      "learning_rate": 4.783783783783784e-06,
+      "loss": 0.3744,
+      "step": 52110
+    },
+    {
+      "epoch": 156.52,
+      "grad_norm": 18.58957862854004,
+      "learning_rate": 4.7827827827827836e-06,
+      "loss": 0.3956,
+      "step": 52120
+    },
+    {
+      "epoch": 156.55,
+      "grad_norm": 17.024513244628906,
+      "learning_rate": 4.781781781781782e-06,
+      "loss": 0.4446,
+      "step": 52130
+    },
+    {
+      "epoch": 156.58,
+      "grad_norm": 10.029541015625,
+      "learning_rate": 4.780780780780781e-06,
+      "loss": 0.3885,
+      "step": 52140
+    },
+    {
+      "epoch": 156.61,
+      "grad_norm": 18.381694793701172,
+      "learning_rate": 4.77977977977978e-06,
+      "loss": 0.3594,
+      "step": 52150
+    },
+    {
+      "epoch": 156.64,
+      "grad_norm": 15.112462043762207,
+      "learning_rate": 4.77877877877878e-06,
+      "loss": 0.3888,
+      "step": 52160
+    },
+    {
+      "epoch": 156.67,
+      "grad_norm": 18.273466110229492,
+      "learning_rate": 4.777777777777778e-06,
+      "loss": 0.3889,
+      "step": 52170
+    },
+    {
+      "epoch": 156.7,
+      "grad_norm": 13.451169967651367,
+      "learning_rate": 4.776776776776777e-06,
+      "loss": 0.3794,
+      "step": 52180
+    },
+    {
+      "epoch": 156.73,
+      "grad_norm": 13.175156593322754,
+      "learning_rate": 4.775775775775776e-06,
+      "loss": 0.3885,
+      "step": 52190
+    },
+    {
+      "epoch": 156.76,
+      "grad_norm": 17.60410499572754,
+      "learning_rate": 4.774774774774775e-06,
+      "loss": 0.3656,
+      "step": 52200
+    },
+    {
+      "epoch": 156.79,
+      "grad_norm": 19.51572608947754,
+      "learning_rate": 4.773773773773774e-06,
+      "loss": 0.3931,
+      "step": 52210
+    },
+    {
+      "epoch": 156.82,
+      "grad_norm": 17.39482307434082,
+      "learning_rate": 4.772772772772773e-06,
+      "loss": 0.3929,
+      "step": 52220
+    },
+    {
+      "epoch": 156.85,
+      "grad_norm": 22.751001358032227,
+      "learning_rate": 4.771771771771772e-06,
+      "loss": 0.4262,
+      "step": 52230
+    },
+    {
+      "epoch": 156.88,
+      "grad_norm": 18.511192321777344,
+      "learning_rate": 4.770770770770771e-06,
+      "loss": 0.4203,
+      "step": 52240
+    },
+    {
+      "epoch": 156.91,
+      "grad_norm": 23.269819259643555,
+      "learning_rate": 4.76976976976977e-06,
+      "loss": 0.4006,
+      "step": 52250
+    },
+    {
+      "epoch": 156.94,
+      "grad_norm": 19.067413330078125,
+      "learning_rate": 4.768768768768769e-06,
+      "loss": 0.4138,
+      "step": 52260
+    },
+    {
+      "epoch": 156.97,
+      "grad_norm": 16.199256896972656,
+      "learning_rate": 4.7677677677677684e-06,
+      "loss": 0.3831,
+      "step": 52270
+    },
+    {
+      "epoch": 157.0,
+      "grad_norm": 15.092903137207031,
+      "learning_rate": 4.766766766766767e-06,
+      "loss": 0.4191,
+      "step": 52280
+    },
+    {
+      "epoch": 157.0,
+      "eval_accuracy": 0.866,
+      "eval_loss": 0.5213327407836914,
+      "eval_runtime": 12.8287,
+      "eval_samples_per_second": 779.503,
+      "eval_steps_per_second": 3.118,
+      "step": 52281
+    },
+    {
+      "epoch": 157.03,
+      "grad_norm": 13.513900756835938,
+      "learning_rate": 4.7657657657657665e-06,
+      "loss": 0.3331,
+      "step": 52290
+    },
+    {
+      "epoch": 157.06,
+      "grad_norm": 13.491900444030762,
+      "learning_rate": 4.764764764764765e-06,
+      "loss": 0.3457,
+      "step": 52300
+    },
+    {
+      "epoch": 157.09,
+      "grad_norm": 20.068584442138672,
+      "learning_rate": 4.763763763763764e-06,
+      "loss": 0.3592,
+      "step": 52310
+    },
+    {
+      "epoch": 157.12,
+      "grad_norm": 19.521589279174805,
+      "learning_rate": 4.762762762762763e-06,
+      "loss": 0.4162,
+      "step": 52320
+    },
+    {
+      "epoch": 157.15,
+      "grad_norm": 26.861099243164062,
+      "learning_rate": 4.7617617617617625e-06,
+      "loss": 0.4082,
+      "step": 52330
+    },
+    {
+      "epoch": 157.18,
+      "grad_norm": 11.813328742980957,
+      "learning_rate": 4.760760760760761e-06,
+      "loss": 0.3961,
+      "step": 52340
+    },
+    {
+      "epoch": 157.21,
+      "grad_norm": 13.858960151672363,
+      "learning_rate": 4.75975975975976e-06,
+      "loss": 0.3956,
+      "step": 52350
+    },
+    {
+      "epoch": 157.24,
+      "grad_norm": 12.994694709777832,
+      "learning_rate": 4.758758758758759e-06,
+      "loss": 0.3935,
+      "step": 52360
+    },
+    {
+      "epoch": 157.27,
+      "grad_norm": 17.658119201660156,
+      "learning_rate": 4.757757757757759e-06,
+      "loss": 0.3759,
+      "step": 52370
+    },
+    {
+      "epoch": 157.3,
+      "grad_norm": 17.56904411315918,
+      "learning_rate": 4.756756756756757e-06,
+      "loss": 0.382,
+      "step": 52380
+    },
+    {
+      "epoch": 157.33,
+      "grad_norm": 15.622025489807129,
+      "learning_rate": 4.755755755755756e-06,
+      "loss": 0.3907,
+      "step": 52390
+    },
+    {
+      "epoch": 157.36,
+      "grad_norm": 16.552410125732422,
+      "learning_rate": 4.754754754754755e-06,
+      "loss": 0.3697,
+      "step": 52400
+    },
+    {
+      "epoch": 157.39,
+      "grad_norm": 17.084484100341797,
+      "learning_rate": 4.753753753753754e-06,
+      "loss": 0.4224,
+      "step": 52410
+    },
+    {
+      "epoch": 157.42,
+      "grad_norm": 12.868242263793945,
+      "learning_rate": 4.752752752752753e-06,
+      "loss": 0.4159,
+      "step": 52420
+    },
+    {
+      "epoch": 157.45,
+      "grad_norm": 23.756208419799805,
+      "learning_rate": 4.751751751751752e-06,
+      "loss": 0.4166,
+      "step": 52430
+    },
+    {
+      "epoch": 157.48,
+      "grad_norm": 18.55318260192871,
+      "learning_rate": 4.750750750750751e-06,
+      "loss": 0.4109,
+      "step": 52440
+    },
+    {
+      "epoch": 157.51,
+      "grad_norm": 16.046588897705078,
+      "learning_rate": 4.74974974974975e-06,
+      "loss": 0.4052,
+      "step": 52450
+    },
+    {
+      "epoch": 157.54,
+      "grad_norm": 18.576801300048828,
+      "learning_rate": 4.748748748748749e-06,
+      "loss": 0.3855,
+      "step": 52460
+    },
+    {
+      "epoch": 157.57,
+      "grad_norm": 14.581674575805664,
+      "learning_rate": 4.747747747747748e-06,
+      "loss": 0.3567,
+      "step": 52470
+    },
+    {
+      "epoch": 157.6,
+      "grad_norm": 14.218754768371582,
+      "learning_rate": 4.746746746746747e-06,
+      "loss": 0.3808,
+      "step": 52480
+    },
+    {
+      "epoch": 157.63,
+      "grad_norm": 12.55324649810791,
+      "learning_rate": 4.745745745745746e-06,
+      "loss": 0.3453,
+      "step": 52490
+    },
+    {
+      "epoch": 157.66,
+      "grad_norm": 15.872807502746582,
+      "learning_rate": 4.7447447447447454e-06,
+      "loss": 0.3961,
+      "step": 52500
+    },
+    {
+      "epoch": 157.69,
+      "grad_norm": 21.042171478271484,
+      "learning_rate": 4.743743743743744e-06,
+      "loss": 0.3999,
+      "step": 52510
+    },
+    {
+      "epoch": 157.72,
+      "grad_norm": 9.493179321289062,
+      "learning_rate": 4.7427427427427435e-06,
+      "loss": 0.3967,
+      "step": 52520
+    },
+    {
+      "epoch": 157.75,
+      "grad_norm": 13.818670272827148,
+      "learning_rate": 4.741741741741742e-06,
+      "loss": 0.3763,
+      "step": 52530
+    },
+    {
+      "epoch": 157.78,
+      "grad_norm": 12.930792808532715,
+      "learning_rate": 4.7407407407407415e-06,
+      "loss": 0.3858,
+      "step": 52540
+    },
+    {
+      "epoch": 157.81,
+      "grad_norm": 19.136394500732422,
+      "learning_rate": 4.73973973973974e-06,
+      "loss": 0.4178,
+      "step": 52550
+    },
+    {
+      "epoch": 157.84,
+      "grad_norm": 14.42113971710205,
+      "learning_rate": 4.738738738738739e-06,
+      "loss": 0.3695,
+      "step": 52560
+    },
+    {
+      "epoch": 157.87,
+      "grad_norm": 21.43897247314453,
+      "learning_rate": 4.737737737737738e-06,
+      "loss": 0.3624,
+      "step": 52570
+    },
+    {
+      "epoch": 157.9,
+      "grad_norm": 21.472152709960938,
+      "learning_rate": 4.7367367367367376e-06,
+      "loss": 0.393,
+      "step": 52580
+    },
+    {
+      "epoch": 157.93,
+      "grad_norm": 11.864038467407227,
+      "learning_rate": 4.735735735735736e-06,
+      "loss": 0.3995,
+      "step": 52590
+    },
+    {
+      "epoch": 157.96,
+      "grad_norm": 13.976758003234863,
+      "learning_rate": 4.734734734734735e-06,
+      "loss": 0.3869,
+      "step": 52600
+    },
+    {
+      "epoch": 157.99,
+      "grad_norm": 12.832066535949707,
+      "learning_rate": 4.733733733733734e-06,
+      "loss": 0.3864,
+      "step": 52610
+    },
+    {
+      "epoch": 158.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.5225269794464111,
+      "eval_runtime": 12.8481,
+      "eval_samples_per_second": 778.327,
+      "eval_steps_per_second": 3.113,
+      "step": 52614
+    },
+    {
+      "epoch": 158.02,
+      "grad_norm": 14.946742057800293,
+      "learning_rate": 4.732732732732734e-06,
+      "loss": 0.4272,
+      "step": 52620
+    },
+    {
+      "epoch": 158.05,
+      "grad_norm": 18.53807258605957,
+      "learning_rate": 4.731731731731732e-06,
+      "loss": 0.4207,
+      "step": 52630
+    },
+    {
+      "epoch": 158.08,
+      "grad_norm": 19.787860870361328,
+      "learning_rate": 4.730730730730731e-06,
+      "loss": 0.3954,
+      "step": 52640
+    },
+    {
+      "epoch": 158.11,
+      "grad_norm": 9.429976463317871,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.3518,
+      "step": 52650
+    },
+    {
+      "epoch": 158.14,
+      "grad_norm": 15.09274673461914,
+      "learning_rate": 4.728728728728729e-06,
+      "loss": 0.3464,
+      "step": 52660
+    },
+    {
+      "epoch": 158.17,
+      "grad_norm": 14.510457992553711,
+      "learning_rate": 4.727727727727728e-06,
+      "loss": 0.3933,
+      "step": 52670
+    },
+    {
+      "epoch": 158.2,
+      "grad_norm": 13.177496910095215,
+      "learning_rate": 4.726726726726727e-06,
+      "loss": 0.3715,
+      "step": 52680
+    },
+    {
+      "epoch": 158.23,
+      "grad_norm": 18.615314483642578,
+      "learning_rate": 4.725725725725726e-06,
+      "loss": 0.3795,
+      "step": 52690
+    },
+    {
+      "epoch": 158.26,
+      "grad_norm": 21.882566452026367,
+      "learning_rate": 4.724724724724725e-06,
+      "loss": 0.3925,
+      "step": 52700
+    },
+    {
+      "epoch": 158.29,
+      "grad_norm": 13.371017456054688,
+      "learning_rate": 4.723723723723724e-06,
+      "loss": 0.4005,
+      "step": 52710
+    },
+    {
+      "epoch": 158.32,
+      "grad_norm": 20.007226943969727,
+      "learning_rate": 4.722722722722723e-06,
+      "loss": 0.4524,
+      "step": 52720
+    },
+    {
+      "epoch": 158.35,
+      "grad_norm": 19.687694549560547,
+      "learning_rate": 4.7217217217217224e-06,
+      "loss": 0.377,
+      "step": 52730
+    },
+    {
+      "epoch": 158.38,
+      "grad_norm": 9.997729301452637,
+      "learning_rate": 4.720720720720721e-06,
+      "loss": 0.3538,
+      "step": 52740
+    },
+    {
+      "epoch": 158.41,
+      "grad_norm": 16.377777099609375,
+      "learning_rate": 4.7197197197197205e-06,
+      "loss": 0.382,
+      "step": 52750
+    },
+    {
+      "epoch": 158.44,
+      "grad_norm": 11.147761344909668,
+      "learning_rate": 4.718718718718719e-06,
+      "loss": 0.4311,
+      "step": 52760
+    },
+    {
+      "epoch": 158.47,
+      "grad_norm": 17.74378204345703,
+      "learning_rate": 4.717717717717718e-06,
+      "loss": 0.4256,
+      "step": 52770
+    },
+    {
+      "epoch": 158.5,
+      "grad_norm": 15.476859092712402,
+      "learning_rate": 4.716716716716717e-06,
+      "loss": 0.4138,
+      "step": 52780
+    },
+    {
+      "epoch": 158.53,
+      "grad_norm": 16.54972267150879,
+      "learning_rate": 4.7157157157157165e-06,
+      "loss": 0.3291,
+      "step": 52790
+    },
+    {
+      "epoch": 158.56,
+      "grad_norm": 15.643482208251953,
+      "learning_rate": 4.714714714714715e-06,
+      "loss": 0.3598,
+      "step": 52800
+    },
+    {
+      "epoch": 158.59,
+      "grad_norm": 25.990219116210938,
+      "learning_rate": 4.713713713713714e-06,
+      "loss": 0.4253,
+      "step": 52810
+    },
+    {
+      "epoch": 158.62,
+      "grad_norm": 11.986838340759277,
+      "learning_rate": 4.712712712712713e-06,
+      "loss": 0.3534,
+      "step": 52820
+    },
+    {
+      "epoch": 158.65,
+      "grad_norm": 17.522445678710938,
+      "learning_rate": 4.711711711711713e-06,
+      "loss": 0.3503,
+      "step": 52830
+    },
+    {
+      "epoch": 158.68,
+      "grad_norm": 17.72920799255371,
+      "learning_rate": 4.710710710710711e-06,
+      "loss": 0.3602,
+      "step": 52840
+    },
+    {
+      "epoch": 158.71,
+      "grad_norm": 25.97089385986328,
+      "learning_rate": 4.70970970970971e-06,
+      "loss": 0.436,
+      "step": 52850
+    },
+    {
+      "epoch": 158.74,
+      "grad_norm": 9.859343528747559,
+      "learning_rate": 4.708708708708709e-06,
+      "loss": 0.3771,
+      "step": 52860
+    },
+    {
+      "epoch": 158.77,
+      "grad_norm": 15.989078521728516,
+      "learning_rate": 4.707707707707708e-06,
+      "loss": 0.4028,
+      "step": 52870
+    },
+    {
+      "epoch": 158.8,
+      "grad_norm": 17.197994232177734,
+      "learning_rate": 4.7067067067067064e-06,
+      "loss": 0.4119,
+      "step": 52880
+    },
+    {
+      "epoch": 158.83,
+      "grad_norm": 12.681829452514648,
+      "learning_rate": 4.705705705705706e-06,
+      "loss": 0.391,
+      "step": 52890
+    },
+    {
+      "epoch": 158.86,
+      "grad_norm": 16.960227966308594,
+      "learning_rate": 4.704704704704705e-06,
+      "loss": 0.3552,
+      "step": 52900
+    },
+    {
+      "epoch": 158.89,
+      "grad_norm": 17.572105407714844,
+      "learning_rate": 4.703703703703704e-06,
+      "loss": 0.3843,
+      "step": 52910
+    },
+    {
+      "epoch": 158.92,
+      "grad_norm": 16.440229415893555,
+      "learning_rate": 4.702702702702703e-06,
+      "loss": 0.3896,
+      "step": 52920
+    },
+    {
+      "epoch": 158.95,
+      "grad_norm": 16.41837501525879,
+      "learning_rate": 4.701701701701702e-06,
+      "loss": 0.3817,
+      "step": 52930
+    },
+    {
+      "epoch": 158.98,
+      "grad_norm": 26.96352195739746,
+      "learning_rate": 4.700700700700701e-06,
+      "loss": 0.3974,
+      "step": 52940
+    },
+    {
+      "epoch": 159.0,
+      "eval_accuracy": 0.8653,
+      "eval_loss": 0.5248140692710876,
+      "eval_runtime": 12.5118,
+      "eval_samples_per_second": 799.248,
+      "eval_steps_per_second": 3.197,
+      "step": 52947
+    },
+    {
+      "epoch": 159.01,
+      "grad_norm": 15.826425552368164,
+      "learning_rate": 4.6996996996997e-06,
+      "loss": 0.3551,
+      "step": 52950
+    },
+    {
+      "epoch": 159.04,
+      "grad_norm": 9.157735824584961,
+      "learning_rate": 4.698698698698699e-06,
+      "loss": 0.3697,
+      "step": 52960
+    },
+    {
+      "epoch": 159.07,
+      "grad_norm": 12.635099411010742,
+      "learning_rate": 4.697697697697698e-06,
+      "loss": 0.4169,
+      "step": 52970
+    },
+    {
+      "epoch": 159.1,
+      "grad_norm": 15.64363956451416,
+      "learning_rate": 4.696696696696697e-06,
+      "loss": 0.4026,
+      "step": 52980
+    },
+    {
+      "epoch": 159.13,
+      "grad_norm": 16.821754455566406,
+      "learning_rate": 4.695695695695696e-06,
+      "loss": 0.4244,
+      "step": 52990
+    },
+    {
+      "epoch": 159.16,
+      "grad_norm": 16.658946990966797,
+      "learning_rate": 4.6946946946946955e-06,
+      "loss": 0.3601,
+      "step": 53000
+    },
+    {
+      "epoch": 159.19,
+      "grad_norm": 19.141950607299805,
+      "learning_rate": 4.693693693693694e-06,
+      "loss": 0.371,
+      "step": 53010
+    },
+    {
+      "epoch": 159.22,
+      "grad_norm": 10.895539283752441,
+      "learning_rate": 4.692692692692693e-06,
+      "loss": 0.3829,
+      "step": 53020
+    },
+    {
+      "epoch": 159.25,
+      "grad_norm": 12.44287109375,
+      "learning_rate": 4.691691691691692e-06,
+      "loss": 0.3546,
+      "step": 53030
+    },
+    {
+      "epoch": 159.28,
+      "grad_norm": 13.975525856018066,
+      "learning_rate": 4.6906906906906916e-06,
+      "loss": 0.4104,
+      "step": 53040
+    },
+    {
+      "epoch": 159.31,
+      "grad_norm": 22.995363235473633,
+      "learning_rate": 4.68968968968969e-06,
+      "loss": 0.3996,
+      "step": 53050
+    },
+    {
+      "epoch": 159.34,
+      "grad_norm": 15.043763160705566,
+      "learning_rate": 4.688688688688689e-06,
+      "loss": 0.3995,
+      "step": 53060
+    },
+    {
+      "epoch": 159.37,
+      "grad_norm": 13.063130378723145,
+      "learning_rate": 4.687687687687688e-06,
+      "loss": 0.3867,
+      "step": 53070
+    },
+    {
+      "epoch": 159.4,
+      "grad_norm": 12.215137481689453,
+      "learning_rate": 4.686686686686688e-06,
+      "loss": 0.3663,
+      "step": 53080
+    },
+    {
+      "epoch": 159.43,
+      "grad_norm": 19.004924774169922,
+      "learning_rate": 4.685685685685686e-06,
+      "loss": 0.4109,
+      "step": 53090
+    },
+    {
+      "epoch": 159.46,
+      "grad_norm": 15.92647933959961,
+      "learning_rate": 4.684684684684685e-06,
+      "loss": 0.3625,
+      "step": 53100
+    },
+    {
+      "epoch": 159.49,
+      "grad_norm": 15.877604484558105,
+      "learning_rate": 4.683683683683684e-06,
+      "loss": 0.3775,
+      "step": 53110
+    },
+    {
+      "epoch": 159.52,
+      "grad_norm": 24.06257438659668,
+      "learning_rate": 4.682682682682683e-06,
+      "loss": 0.4168,
+      "step": 53120
+    },
+    {
+      "epoch": 159.55,
+      "grad_norm": 11.87307357788086,
+      "learning_rate": 4.6816816816816815e-06,
+      "loss": 0.4338,
+      "step": 53130
+    },
+    {
+      "epoch": 159.58,
+      "grad_norm": 11.275247573852539,
+      "learning_rate": 4.680680680680681e-06,
+      "loss": 0.3402,
+      "step": 53140
+    },
+    {
+      "epoch": 159.61,
+      "grad_norm": 18.39348793029785,
+      "learning_rate": 4.67967967967968e-06,
+      "loss": 0.3712,
+      "step": 53150
+    },
+    {
+      "epoch": 159.64,
+      "grad_norm": 18.498178482055664,
+      "learning_rate": 4.678678678678679e-06,
+      "loss": 0.4108,
+      "step": 53160
+    },
+    {
+      "epoch": 159.67,
+      "grad_norm": 13.822325706481934,
+      "learning_rate": 4.677677677677678e-06,
+      "loss": 0.4238,
+      "step": 53170
+    },
+    {
+      "epoch": 159.7,
+      "grad_norm": 12.693727493286133,
+      "learning_rate": 4.676676676676677e-06,
+      "loss": 0.3868,
+      "step": 53180
+    },
+    {
+      "epoch": 159.73,
+      "grad_norm": 22.476547241210938,
+      "learning_rate": 4.675675675675676e-06,
+      "loss": 0.4101,
+      "step": 53190
+    },
+    {
+      "epoch": 159.76,
+      "grad_norm": 21.43368911743164,
+      "learning_rate": 4.674674674674675e-06,
+      "loss": 0.3713,
+      "step": 53200
+    },
+    {
+      "epoch": 159.79,
+      "grad_norm": 9.922982215881348,
+      "learning_rate": 4.6736736736736745e-06,
+      "loss": 0.4364,
+      "step": 53210
+    },
+    {
+      "epoch": 159.82,
+      "grad_norm": 16.925371170043945,
+      "learning_rate": 4.672672672672673e-06,
+      "loss": 0.3662,
+      "step": 53220
+    },
+    {
+      "epoch": 159.85,
+      "grad_norm": 16.639142990112305,
+      "learning_rate": 4.671671671671672e-06,
+      "loss": 0.3873,
+      "step": 53230
+    },
+    {
+      "epoch": 159.88,
+      "grad_norm": 10.712470054626465,
+      "learning_rate": 4.670670670670671e-06,
+      "loss": 0.4264,
+      "step": 53240
+    },
+    {
+      "epoch": 159.91,
+      "grad_norm": 9.30695915222168,
+      "learning_rate": 4.6696696696696705e-06,
+      "loss": 0.4053,
+      "step": 53250
+    },
+    {
+      "epoch": 159.94,
+      "grad_norm": 8.743491172790527,
+      "learning_rate": 4.668668668668669e-06,
+      "loss": 0.3553,
+      "step": 53260
+    },
+    {
+      "epoch": 159.97,
+      "grad_norm": 14.263792991638184,
+      "learning_rate": 4.667667667667668e-06,
+      "loss": 0.4042,
+      "step": 53270
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 18.900970458984375,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.355,
+      "step": 53280
+    },
+    {
+      "epoch": 160.0,
+      "eval_accuracy": 0.8626,
+      "eval_loss": 0.5264528393745422,
+      "eval_runtime": 12.8188,
+      "eval_samples_per_second": 780.107,
+      "eval_steps_per_second": 3.12,
+      "step": 53280
+    },
+    {
+      "epoch": 160.03,
+      "grad_norm": 12.903138160705566,
+      "learning_rate": 4.665665665665667e-06,
+      "loss": 0.3656,
+      "step": 53290
+    },
+    {
+      "epoch": 160.06,
+      "grad_norm": 29.77675437927246,
+      "learning_rate": 4.664664664664665e-06,
+      "loss": 0.4057,
+      "step": 53300
+    },
+    {
+      "epoch": 160.09,
+      "grad_norm": 21.246644973754883,
+      "learning_rate": 4.663663663663664e-06,
+      "loss": 0.384,
+      "step": 53310
+    },
+    {
+      "epoch": 160.12,
+      "grad_norm": 17.53751564025879,
+      "learning_rate": 4.662662662662663e-06,
+      "loss": 0.3318,
+      "step": 53320
+    },
+    {
+      "epoch": 160.15,
+      "grad_norm": 25.563364028930664,
+      "learning_rate": 4.661661661661662e-06,
+      "loss": 0.4009,
+      "step": 53330
+    },
+    {
+      "epoch": 160.18,
+      "grad_norm": 11.281298637390137,
+      "learning_rate": 4.6606606606606604e-06,
+      "loss": 0.4183,
+      "step": 53340
+    },
+    {
+      "epoch": 160.21,
+      "grad_norm": 13.018889427185059,
+      "learning_rate": 4.65965965965966e-06,
+      "loss": 0.4282,
+      "step": 53350
+    },
+    {
+      "epoch": 160.24,
+      "grad_norm": 17.679872512817383,
+      "learning_rate": 4.658658658658659e-06,
+      "loss": 0.3648,
+      "step": 53360
+    },
+    {
+      "epoch": 160.27,
+      "grad_norm": 13.100308418273926,
+      "learning_rate": 4.657657657657658e-06,
+      "loss": 0.4212,
+      "step": 53370
+    },
+    {
+      "epoch": 160.3,
+      "grad_norm": 17.66975212097168,
+      "learning_rate": 4.6566566566566565e-06,
+      "loss": 0.382,
+      "step": 53380
+    },
+    {
+      "epoch": 160.33,
+      "grad_norm": 20.252641677856445,
+      "learning_rate": 4.655655655655656e-06,
+      "loss": 0.4217,
+      "step": 53390
+    },
+    {
+      "epoch": 160.36,
+      "grad_norm": 20.051387786865234,
+      "learning_rate": 4.654654654654655e-06,
+      "loss": 0.3675,
+      "step": 53400
+    },
+    {
+      "epoch": 160.39,
+      "grad_norm": 11.914758682250977,
+      "learning_rate": 4.653653653653654e-06,
+      "loss": 0.3794,
+      "step": 53410
+    },
+    {
+      "epoch": 160.42,
+      "grad_norm": 16.052852630615234,
+      "learning_rate": 4.6526526526526526e-06,
+      "loss": 0.3763,
+      "step": 53420
+    },
+    {
+      "epoch": 160.45,
+      "grad_norm": 35.3156852722168,
+      "learning_rate": 4.651651651651652e-06,
+      "loss": 0.3746,
+      "step": 53430
+    },
+    {
+      "epoch": 160.48,
+      "grad_norm": 18.11252212524414,
+      "learning_rate": 4.650650650650651e-06,
+      "loss": 0.3554,
+      "step": 53440
+    },
+    {
+      "epoch": 160.51,
+      "grad_norm": 12.814082145690918,
+      "learning_rate": 4.64964964964965e-06,
+      "loss": 0.4052,
+      "step": 53450
+    },
+    {
+      "epoch": 160.54,
+      "grad_norm": 12.778681755065918,
+      "learning_rate": 4.6486486486486495e-06,
+      "loss": 0.3965,
+      "step": 53460
+    },
+    {
+      "epoch": 160.57,
+      "grad_norm": 25.69891357421875,
+      "learning_rate": 4.647647647647648e-06,
+      "loss": 0.355,
+      "step": 53470
+    },
+    {
+      "epoch": 160.6,
+      "grad_norm": 13.406737327575684,
+      "learning_rate": 4.646646646646647e-06,
+      "loss": 0.3451,
+      "step": 53480
+    },
+    {
+      "epoch": 160.63,
+      "grad_norm": 30.708301544189453,
+      "learning_rate": 4.645645645645646e-06,
+      "loss": 0.4136,
+      "step": 53490
+    },
+    {
+      "epoch": 160.66,
+      "grad_norm": 17.802513122558594,
+      "learning_rate": 4.6446446446446456e-06,
+      "loss": 0.3498,
+      "step": 53500
+    },
+    {
+      "epoch": 160.69,
+      "grad_norm": 18.531147003173828,
+      "learning_rate": 4.643643643643644e-06,
+      "loss": 0.381,
+      "step": 53510
+    },
+    {
+      "epoch": 160.72,
+      "grad_norm": 20.2080135345459,
+      "learning_rate": 4.642642642642643e-06,
+      "loss": 0.3777,
+      "step": 53520
+    },
+    {
+      "epoch": 160.75,
+      "grad_norm": 17.559789657592773,
+      "learning_rate": 4.641641641641642e-06,
+      "loss": 0.3741,
+      "step": 53530
+    },
+    {
+      "epoch": 160.78,
+      "grad_norm": 19.440393447875977,
+      "learning_rate": 4.640640640640642e-06,
+      "loss": 0.3816,
+      "step": 53540
+    },
+    {
+      "epoch": 160.81,
+      "grad_norm": 16.53251838684082,
+      "learning_rate": 4.63963963963964e-06,
+      "loss": 0.4028,
+      "step": 53550
+    },
+    {
+      "epoch": 160.84,
+      "grad_norm": 14.898193359375,
+      "learning_rate": 4.638638638638639e-06,
+      "loss": 0.3769,
+      "step": 53560
+    },
+    {
+      "epoch": 160.87,
+      "grad_norm": 37.31252670288086,
+      "learning_rate": 4.637637637637638e-06,
+      "loss": 0.3898,
+      "step": 53570
+    },
+    {
+      "epoch": 160.9,
+      "grad_norm": 13.81872272491455,
+      "learning_rate": 4.636636636636637e-06,
+      "loss": 0.4856,
+      "step": 53580
+    },
+    {
+      "epoch": 160.93,
+      "grad_norm": 8.692588806152344,
+      "learning_rate": 4.6356356356356355e-06,
+      "loss": 0.3726,
+      "step": 53590
+    },
+    {
+      "epoch": 160.96,
+      "grad_norm": 13.435314178466797,
+      "learning_rate": 4.634634634634635e-06,
+      "loss": 0.3654,
+      "step": 53600
+    },
+    {
+      "epoch": 160.99,
+      "grad_norm": 18.67369842529297,
+      "learning_rate": 4.633633633633634e-06,
+      "loss": 0.3511,
+      "step": 53610
+    },
+    {
+      "epoch": 161.0,
+      "eval_accuracy": 0.8665,
+      "eval_loss": 0.5227054357528687,
+      "eval_runtime": 12.4832,
+      "eval_samples_per_second": 801.079,
+      "eval_steps_per_second": 3.204,
+      "step": 53613
+    },
+    {
+      "epoch": 161.02,
+      "grad_norm": 18.05752182006836,
+      "learning_rate": 4.632632632632633e-06,
+      "loss": 0.314,
+      "step": 53620
+    },
+    {
+      "epoch": 161.05,
+      "grad_norm": 13.95981216430664,
+      "learning_rate": 4.6316316316316315e-06,
+      "loss": 0.3735,
+      "step": 53630
+    },
+    {
+      "epoch": 161.08,
+      "grad_norm": 14.502241134643555,
+      "learning_rate": 4.630630630630631e-06,
+      "loss": 0.4393,
+      "step": 53640
+    },
+    {
+      "epoch": 161.11,
+      "grad_norm": 11.138081550598145,
+      "learning_rate": 4.62962962962963e-06,
+      "loss": 0.395,
+      "step": 53650
+    },
+    {
+      "epoch": 161.14,
+      "grad_norm": 14.196554183959961,
+      "learning_rate": 4.628628628628629e-06,
+      "loss": 0.3863,
+      "step": 53660
+    },
+    {
+      "epoch": 161.17,
+      "grad_norm": 16.387664794921875,
+      "learning_rate": 4.627627627627628e-06,
+      "loss": 0.41,
+      "step": 53670
+    },
+    {
+      "epoch": 161.2,
+      "grad_norm": 22.404287338256836,
+      "learning_rate": 4.626626626626627e-06,
+      "loss": 0.3398,
+      "step": 53680
+    },
+    {
+      "epoch": 161.23,
+      "grad_norm": 19.634933471679688,
+      "learning_rate": 4.625625625625626e-06,
+      "loss": 0.4071,
+      "step": 53690
+    },
+    {
+      "epoch": 161.26,
+      "grad_norm": 13.614376068115234,
+      "learning_rate": 4.624624624624625e-06,
+      "loss": 0.4291,
+      "step": 53700
+    },
+    {
+      "epoch": 161.29,
+      "grad_norm": 16.84930419921875,
+      "learning_rate": 4.6236236236236245e-06,
+      "loss": 0.3912,
+      "step": 53710
+    },
+    {
+      "epoch": 161.32,
+      "grad_norm": 14.674084663391113,
+      "learning_rate": 4.622622622622623e-06,
+      "loss": 0.4194,
+      "step": 53720
+    },
+    {
+      "epoch": 161.35,
+      "grad_norm": 20.6453800201416,
+      "learning_rate": 4.621621621621622e-06,
+      "loss": 0.3811,
+      "step": 53730
+    },
+    {
+      "epoch": 161.38,
+      "grad_norm": 18.67611312866211,
+      "learning_rate": 4.620620620620621e-06,
+      "loss": 0.4267,
+      "step": 53740
+    },
+    {
+      "epoch": 161.41,
+      "grad_norm": 14.124916076660156,
+      "learning_rate": 4.619619619619621e-06,
+      "loss": 0.3195,
+      "step": 53750
+    },
+    {
+      "epoch": 161.44,
+      "grad_norm": 14.113972663879395,
+      "learning_rate": 4.618618618618619e-06,
+      "loss": 0.4104,
+      "step": 53760
+    },
+    {
+      "epoch": 161.47,
+      "grad_norm": 16.16220474243164,
+      "learning_rate": 4.617617617617618e-06,
+      "loss": 0.4198,
+      "step": 53770
+    },
+    {
+      "epoch": 161.5,
+      "grad_norm": 17.035978317260742,
+      "learning_rate": 4.616616616616617e-06,
+      "loss": 0.3842,
+      "step": 53780
+    },
+    {
+      "epoch": 161.53,
+      "grad_norm": 13.418736457824707,
+      "learning_rate": 4.615615615615616e-06,
+      "loss": 0.3744,
+      "step": 53790
+    },
+    {
+      "epoch": 161.56,
+      "grad_norm": 13.824076652526855,
+      "learning_rate": 4.614614614614614e-06,
+      "loss": 0.3813,
+      "step": 53800
+    },
+    {
+      "epoch": 161.59,
+      "grad_norm": 9.859395027160645,
+      "learning_rate": 4.613613613613614e-06,
+      "loss": 0.4036,
+      "step": 53810
+    },
+    {
+      "epoch": 161.62,
+      "grad_norm": 13.620723724365234,
+      "learning_rate": 4.612612612612613e-06,
+      "loss": 0.3623,
+      "step": 53820
+    },
+    {
+      "epoch": 161.65,
+      "grad_norm": 21.529939651489258,
+      "learning_rate": 4.611611611611612e-06,
+      "loss": 0.383,
+      "step": 53830
+    },
+    {
+      "epoch": 161.68,
+      "grad_norm": 20.136632919311523,
+      "learning_rate": 4.6106106106106105e-06,
+      "loss": 0.3711,
+      "step": 53840
+    },
+    {
+      "epoch": 161.71,
+      "grad_norm": 16.792226791381836,
+      "learning_rate": 4.60960960960961e-06,
+      "loss": 0.4236,
+      "step": 53850
+    },
+    {
+      "epoch": 161.74,
+      "grad_norm": 21.59659767150879,
+      "learning_rate": 4.608608608608609e-06,
+      "loss": 0.34,
+      "step": 53860
+    },
+    {
+      "epoch": 161.77,
+      "grad_norm": 16.25273895263672,
+      "learning_rate": 4.607607607607608e-06,
+      "loss": 0.4242,
+      "step": 53870
+    },
+    {
+      "epoch": 161.8,
+      "grad_norm": 16.098560333251953,
+      "learning_rate": 4.6066066066066066e-06,
+      "loss": 0.4005,
+      "step": 53880
+    },
+    {
+      "epoch": 161.83,
+      "grad_norm": 16.24483299255371,
+      "learning_rate": 4.605605605605606e-06,
+      "loss": 0.4273,
+      "step": 53890
+    },
+    {
+      "epoch": 161.86,
+      "grad_norm": 13.621635437011719,
+      "learning_rate": 4.604604604604605e-06,
+      "loss": 0.3909,
+      "step": 53900
+    },
+    {
+      "epoch": 161.89,
+      "grad_norm": 14.946403503417969,
+      "learning_rate": 4.603603603603604e-06,
+      "loss": 0.382,
+      "step": 53910
+    },
+    {
+      "epoch": 161.92,
+      "grad_norm": 8.650283813476562,
+      "learning_rate": 4.602602602602603e-06,
+      "loss": 0.3637,
+      "step": 53920
+    },
+    {
+      "epoch": 161.95,
+      "grad_norm": 17.3676700592041,
+      "learning_rate": 4.601601601601602e-06,
+      "loss": 0.3653,
+      "step": 53930
+    },
+    {
+      "epoch": 161.98,
+      "grad_norm": 14.457533836364746,
+      "learning_rate": 4.600600600600601e-06,
+      "loss": 0.3945,
+      "step": 53940
+    },
+    {
+      "epoch": 162.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.5200738310813904,
+      "eval_runtime": 13.0164,
+      "eval_samples_per_second": 768.26,
+      "eval_steps_per_second": 3.073,
+      "step": 53946
+    },
+    {
+      "epoch": 162.01,
+      "grad_norm": 17.463367462158203,
+      "learning_rate": 4.5995995995996e-06,
+      "loss": 0.3756,
+      "step": 53950
+    },
+    {
+      "epoch": 162.04,
+      "grad_norm": 16.555734634399414,
+      "learning_rate": 4.5985985985985996e-06,
+      "loss": 0.3762,
+      "step": 53960
+    },
+    {
+      "epoch": 162.07,
+      "grad_norm": 18.439983367919922,
+      "learning_rate": 4.597597597597598e-06,
+      "loss": 0.4073,
+      "step": 53970
+    },
+    {
+      "epoch": 162.1,
+      "grad_norm": 11.80670166015625,
+      "learning_rate": 4.596596596596597e-06,
+      "loss": 0.3857,
+      "step": 53980
+    },
+    {
+      "epoch": 162.13,
+      "grad_norm": 15.096915245056152,
+      "learning_rate": 4.595595595595596e-06,
+      "loss": 0.4443,
+      "step": 53990
+    },
+    {
+      "epoch": 162.16,
+      "grad_norm": 14.283059120178223,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.3817,
+      "step": 54000
+    },
+    {
+      "epoch": 162.19,
+      "grad_norm": 10.434173583984375,
+      "learning_rate": 4.593593593593593e-06,
+      "loss": 0.3935,
+      "step": 54010
+    },
+    {
+      "epoch": 162.22,
+      "grad_norm": 21.55885887145996,
+      "learning_rate": 4.592592592592593e-06,
+      "loss": 0.3497,
+      "step": 54020
+    },
+    {
+      "epoch": 162.25,
+      "grad_norm": 25.191436767578125,
+      "learning_rate": 4.591591591591592e-06,
+      "loss": 0.3488,
+      "step": 54030
+    },
+    {
+      "epoch": 162.28,
+      "grad_norm": 15.999924659729004,
+      "learning_rate": 4.590590590590591e-06,
+      "loss": 0.3616,
+      "step": 54040
+    },
+    {
+      "epoch": 162.31,
+      "grad_norm": 14.4332275390625,
+      "learning_rate": 4.5895895895895895e-06,
+      "loss": 0.3824,
+      "step": 54050
+    },
+    {
+      "epoch": 162.34,
+      "grad_norm": 22.30219268798828,
+      "learning_rate": 4.588588588588589e-06,
+      "loss": 0.4211,
+      "step": 54060
+    },
+    {
+      "epoch": 162.37,
+      "grad_norm": 16.727378845214844,
+      "learning_rate": 4.587587587587588e-06,
+      "loss": 0.3697,
+      "step": 54070
+    },
+    {
+      "epoch": 162.4,
+      "grad_norm": 11.635109901428223,
+      "learning_rate": 4.586586586586587e-06,
+      "loss": 0.4274,
+      "step": 54080
+    },
+    {
+      "epoch": 162.43,
+      "grad_norm": 14.00708293914795,
+      "learning_rate": 4.5855855855855855e-06,
+      "loss": 0.3684,
+      "step": 54090
+    },
+    {
+      "epoch": 162.46,
+      "grad_norm": 20.32889175415039,
+      "learning_rate": 4.584584584584585e-06,
+      "loss": 0.4491,
+      "step": 54100
+    },
+    {
+      "epoch": 162.49,
+      "grad_norm": 23.56399154663086,
+      "learning_rate": 4.583583583583584e-06,
+      "loss": 0.3694,
+      "step": 54110
+    },
+    {
+      "epoch": 162.52,
+      "grad_norm": 16.375730514526367,
+      "learning_rate": 4.582582582582583e-06,
+      "loss": 0.3691,
+      "step": 54120
+    },
+    {
+      "epoch": 162.55,
+      "grad_norm": 11.679108619689941,
+      "learning_rate": 4.581581581581582e-06,
+      "loss": 0.357,
+      "step": 54130
+    },
+    {
+      "epoch": 162.58,
+      "grad_norm": 15.062808990478516,
+      "learning_rate": 4.580580580580581e-06,
+      "loss": 0.3705,
+      "step": 54140
+    },
+    {
+      "epoch": 162.61,
+      "grad_norm": 23.183149337768555,
+      "learning_rate": 4.57957957957958e-06,
+      "loss": 0.4125,
+      "step": 54150
+    },
+    {
+      "epoch": 162.64,
+      "grad_norm": 15.577582359313965,
+      "learning_rate": 4.578578578578579e-06,
+      "loss": 0.4,
+      "step": 54160
+    },
+    {
+      "epoch": 162.67,
+      "grad_norm": 17.682666778564453,
+      "learning_rate": 4.577577577577578e-06,
+      "loss": 0.387,
+      "step": 54170
+    },
+    {
+      "epoch": 162.7,
+      "grad_norm": 22.250558853149414,
+      "learning_rate": 4.576576576576577e-06,
+      "loss": 0.4387,
+      "step": 54180
+    },
+    {
+      "epoch": 162.73,
+      "grad_norm": 16.11025047302246,
+      "learning_rate": 4.575575575575576e-06,
+      "loss": 0.4077,
+      "step": 54190
+    },
+    {
+      "epoch": 162.76,
+      "grad_norm": 9.974811553955078,
+      "learning_rate": 4.574574574574575e-06,
+      "loss": 0.3691,
+      "step": 54200
+    },
+    {
+      "epoch": 162.79,
+      "grad_norm": 19.652097702026367,
+      "learning_rate": 4.573573573573574e-06,
+      "loss": 0.4025,
+      "step": 54210
+    },
+    {
+      "epoch": 162.82,
+      "grad_norm": 18.105653762817383,
+      "learning_rate": 4.572572572572573e-06,
+      "loss": 0.3631,
+      "step": 54220
+    },
+    {
+      "epoch": 162.85,
+      "grad_norm": 15.436280250549316,
+      "learning_rate": 4.571571571571572e-06,
+      "loss": 0.3404,
+      "step": 54230
+    },
+    {
+      "epoch": 162.88,
+      "grad_norm": 14.030817031860352,
+      "learning_rate": 4.570570570570571e-06,
+      "loss": 0.4145,
+      "step": 54240
+    },
+    {
+      "epoch": 162.91,
+      "grad_norm": 13.15462875366211,
+      "learning_rate": 4.56956956956957e-06,
+      "loss": 0.3713,
+      "step": 54250
+    },
+    {
+      "epoch": 162.94,
+      "grad_norm": 18.54828643798828,
+      "learning_rate": 4.568568568568568e-06,
+      "loss": 0.403,
+      "step": 54260
+    },
+    {
+      "epoch": 162.97,
+      "grad_norm": 17.173538208007812,
+      "learning_rate": 4.567567567567568e-06,
+      "loss": 0.3869,
+      "step": 54270
+    },
+    {
+      "epoch": 163.0,
+      "eval_accuracy": 0.8633,
+      "eval_loss": 0.5279990434646606,
+      "eval_runtime": 12.9104,
+      "eval_samples_per_second": 774.57,
+      "eval_steps_per_second": 3.098,
+      "step": 54279
+    },
+    {
+      "epoch": 163.0,
+      "grad_norm": 15.60637378692627,
+      "learning_rate": 4.566566566566567e-06,
+      "loss": 0.3495,
+      "step": 54280
+    },
+    {
+      "epoch": 163.03,
+      "grad_norm": 21.99359130859375,
+      "learning_rate": 4.565565565565566e-06,
+      "loss": 0.3877,
+      "step": 54290
+    },
+    {
+      "epoch": 163.06,
+      "grad_norm": 13.232500076293945,
+      "learning_rate": 4.5645645645645645e-06,
+      "loss": 0.3789,
+      "step": 54300
+    },
+    {
+      "epoch": 163.09,
+      "grad_norm": 20.67872428894043,
+      "learning_rate": 4.563563563563564e-06,
+      "loss": 0.377,
+      "step": 54310
+    },
+    {
+      "epoch": 163.12,
+      "grad_norm": 20.659568786621094,
+      "learning_rate": 4.562562562562563e-06,
+      "loss": 0.42,
+      "step": 54320
+    },
+    {
+      "epoch": 163.15,
+      "grad_norm": 16.62511444091797,
+      "learning_rate": 4.561561561561562e-06,
+      "loss": 0.4272,
+      "step": 54330
+    },
+    {
+      "epoch": 163.18,
+      "grad_norm": 12.446496963500977,
+      "learning_rate": 4.5605605605605606e-06,
+      "loss": 0.4005,
+      "step": 54340
+    },
+    {
+      "epoch": 163.21,
+      "grad_norm": 13.612898826599121,
+      "learning_rate": 4.55955955955956e-06,
+      "loss": 0.4225,
+      "step": 54350
+    },
+    {
+      "epoch": 163.24,
+      "grad_norm": 12.362308502197266,
+      "learning_rate": 4.558558558558559e-06,
+      "loss": 0.3803,
+      "step": 54360
+    },
+    {
+      "epoch": 163.27,
+      "grad_norm": 14.36428451538086,
+      "learning_rate": 4.557557557557558e-06,
+      "loss": 0.3573,
+      "step": 54370
+    },
+    {
+      "epoch": 163.3,
+      "grad_norm": 15.58007526397705,
+      "learning_rate": 4.556556556556557e-06,
+      "loss": 0.3693,
+      "step": 54380
+    },
+    {
+      "epoch": 163.33,
+      "grad_norm": 18.192432403564453,
+      "learning_rate": 4.555555555555556e-06,
+      "loss": 0.3465,
+      "step": 54390
+    },
+    {
+      "epoch": 163.36,
+      "grad_norm": 11.753549575805664,
+      "learning_rate": 4.554554554554555e-06,
+      "loss": 0.4082,
+      "step": 54400
+    },
+    {
+      "epoch": 163.39,
+      "grad_norm": 26.775779724121094,
+      "learning_rate": 4.553553553553554e-06,
+      "loss": 0.3682,
+      "step": 54410
+    },
+    {
+      "epoch": 163.42,
+      "grad_norm": 19.826656341552734,
+      "learning_rate": 4.552552552552553e-06,
+      "loss": 0.3918,
+      "step": 54420
+    },
+    {
+      "epoch": 163.45,
+      "grad_norm": 13.271645545959473,
+      "learning_rate": 4.551551551551552e-06,
+      "loss": 0.3646,
+      "step": 54430
+    },
+    {
+      "epoch": 163.48,
+      "grad_norm": 19.538232803344727,
+      "learning_rate": 4.550550550550551e-06,
+      "loss": 0.3708,
+      "step": 54440
+    },
+    {
+      "epoch": 163.51,
+      "grad_norm": 11.138158798217773,
+      "learning_rate": 4.54954954954955e-06,
+      "loss": 0.4234,
+      "step": 54450
+    },
+    {
+      "epoch": 163.54,
+      "grad_norm": 64.61128234863281,
+      "learning_rate": 4.548548548548549e-06,
+      "loss": 0.4074,
+      "step": 54460
+    },
+    {
+      "epoch": 163.57,
+      "grad_norm": 15.521676063537598,
+      "learning_rate": 4.547547547547547e-06,
+      "loss": 0.3994,
+      "step": 54470
+    },
+    {
+      "epoch": 163.6,
+      "grad_norm": 22.964332580566406,
+      "learning_rate": 4.546546546546547e-06,
+      "loss": 0.3888,
+      "step": 54480
+    },
+    {
+      "epoch": 163.63,
+      "grad_norm": 14.809975624084473,
+      "learning_rate": 4.545545545545546e-06,
+      "loss": 0.3411,
+      "step": 54490
+    },
+    {
+      "epoch": 163.66,
+      "grad_norm": 12.93753433227539,
+      "learning_rate": 4.544544544544545e-06,
+      "loss": 0.3045,
+      "step": 54500
+    },
+    {
+      "epoch": 163.69,
+      "grad_norm": 16.447046279907227,
+      "learning_rate": 4.5435435435435435e-06,
+      "loss": 0.38,
+      "step": 54510
+    },
+    {
+      "epoch": 163.72,
+      "grad_norm": 13.614754676818848,
+      "learning_rate": 4.542542542542543e-06,
+      "loss": 0.3453,
+      "step": 54520
+    },
+    {
+      "epoch": 163.75,
+      "grad_norm": 13.85403060913086,
+      "learning_rate": 4.541541541541542e-06,
+      "loss": 0.3478,
+      "step": 54530
+    },
+    {
+      "epoch": 163.78,
+      "grad_norm": 9.51019287109375,
+      "learning_rate": 4.540540540540541e-06,
+      "loss": 0.3379,
+      "step": 54540
+    },
+    {
+      "epoch": 163.81,
+      "grad_norm": 14.631026268005371,
+      "learning_rate": 4.5395395395395395e-06,
+      "loss": 0.3825,
+      "step": 54550
+    },
+    {
+      "epoch": 163.84,
+      "grad_norm": 16.611286163330078,
+      "learning_rate": 4.538538538538539e-06,
+      "loss": 0.3857,
+      "step": 54560
+    },
+    {
+      "epoch": 163.87,
+      "grad_norm": 17.029569625854492,
+      "learning_rate": 4.537537537537538e-06,
+      "loss": 0.3447,
+      "step": 54570
+    },
+    {
+      "epoch": 163.9,
+      "grad_norm": 18.5035343170166,
+      "learning_rate": 4.536536536536537e-06,
+      "loss": 0.4265,
+      "step": 54580
+    },
+    {
+      "epoch": 163.93,
+      "grad_norm": 22.432737350463867,
+      "learning_rate": 4.535535535535536e-06,
+      "loss": 0.4235,
+      "step": 54590
+    },
+    {
+      "epoch": 163.96,
+      "grad_norm": 12.715253829956055,
+      "learning_rate": 4.534534534534535e-06,
+      "loss": 0.3794,
+      "step": 54600
+    },
+    {
+      "epoch": 163.99,
+      "grad_norm": 34.40452575683594,
+      "learning_rate": 4.533533533533534e-06,
+      "loss": 0.4148,
+      "step": 54610
+    },
+    {
+      "epoch": 164.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5258116722106934,
+      "eval_runtime": 12.9737,
+      "eval_samples_per_second": 770.788,
+      "eval_steps_per_second": 3.083,
+      "step": 54612
+    },
+    {
+      "epoch": 164.02,
+      "grad_norm": 31.696044921875,
+      "learning_rate": 4.532532532532533e-06,
+      "loss": 0.3866,
+      "step": 54620
+    },
+    {
+      "epoch": 164.05,
+      "grad_norm": 14.355303764343262,
+      "learning_rate": 4.531531531531532e-06,
+      "loss": 0.4016,
+      "step": 54630
+    },
+    {
+      "epoch": 164.08,
+      "grad_norm": 9.988592147827148,
+      "learning_rate": 4.530530530530531e-06,
+      "loss": 0.3948,
+      "step": 54640
+    },
+    {
+      "epoch": 164.11,
+      "grad_norm": 12.965171813964844,
+      "learning_rate": 4.52952952952953e-06,
+      "loss": 0.3836,
+      "step": 54650
+    },
+    {
+      "epoch": 164.14,
+      "grad_norm": 14.151989936828613,
+      "learning_rate": 4.528528528528529e-06,
+      "loss": 0.4116,
+      "step": 54660
+    },
+    {
+      "epoch": 164.17,
+      "grad_norm": 10.8665132522583,
+      "learning_rate": 4.527527527527528e-06,
+      "loss": 0.3881,
+      "step": 54670
+    },
+    {
+      "epoch": 164.2,
+      "grad_norm": 17.303848266601562,
+      "learning_rate": 4.526526526526527e-06,
+      "loss": 0.3977,
+      "step": 54680
+    },
+    {
+      "epoch": 164.23,
+      "grad_norm": 18.224245071411133,
+      "learning_rate": 4.525525525525526e-06,
+      "loss": 0.352,
+      "step": 54690
+    },
+    {
+      "epoch": 164.26,
+      "grad_norm": 10.655191421508789,
+      "learning_rate": 4.524524524524525e-06,
+      "loss": 0.3559,
+      "step": 54700
+    },
+    {
+      "epoch": 164.29,
+      "grad_norm": 16.3812255859375,
+      "learning_rate": 4.523523523523524e-06,
+      "loss": 0.3553,
+      "step": 54710
+    },
+    {
+      "epoch": 164.32,
+      "grad_norm": 10.908594131469727,
+      "learning_rate": 4.522522522522522e-06,
+      "loss": 0.3798,
+      "step": 54720
+    },
+    {
+      "epoch": 164.35,
+      "grad_norm": 15.810680389404297,
+      "learning_rate": 4.521521521521522e-06,
+      "loss": 0.3859,
+      "step": 54730
+    },
+    {
+      "epoch": 164.38,
+      "grad_norm": 12.369874000549316,
+      "learning_rate": 4.520520520520521e-06,
+      "loss": 0.4379,
+      "step": 54740
+    },
+    {
+      "epoch": 164.41,
+      "grad_norm": 17.606260299682617,
+      "learning_rate": 4.51951951951952e-06,
+      "loss": 0.333,
+      "step": 54750
+    },
+    {
+      "epoch": 164.44,
+      "grad_norm": 16.25883674621582,
+      "learning_rate": 4.5185185185185185e-06,
+      "loss": 0.3737,
+      "step": 54760
+    },
+    {
+      "epoch": 164.47,
+      "grad_norm": 17.67821502685547,
+      "learning_rate": 4.517517517517518e-06,
+      "loss": 0.3524,
+      "step": 54770
+    },
+    {
+      "epoch": 164.5,
+      "grad_norm": 14.016772270202637,
+      "learning_rate": 4.516516516516517e-06,
+      "loss": 0.3956,
+      "step": 54780
+    },
+    {
+      "epoch": 164.53,
+      "grad_norm": 23.110021591186523,
+      "learning_rate": 4.515515515515516e-06,
+      "loss": 0.3345,
+      "step": 54790
+    },
+    {
+      "epoch": 164.56,
+      "grad_norm": 15.582276344299316,
+      "learning_rate": 4.5145145145145146e-06,
+      "loss": 0.4232,
+      "step": 54800
+    },
+    {
+      "epoch": 164.59,
+      "grad_norm": 18.03217124938965,
+      "learning_rate": 4.513513513513514e-06,
+      "loss": 0.3767,
+      "step": 54810
+    },
+    {
+      "epoch": 164.62,
+      "grad_norm": 15.986672401428223,
+      "learning_rate": 4.512512512512513e-06,
+      "loss": 0.3983,
+      "step": 54820
+    },
+    {
+      "epoch": 164.65,
+      "grad_norm": 13.631553649902344,
+      "learning_rate": 4.511511511511512e-06,
+      "loss": 0.3325,
+      "step": 54830
+    },
+    {
+      "epoch": 164.68,
+      "grad_norm": 13.58836555480957,
+      "learning_rate": 4.510510510510511e-06,
+      "loss": 0.3513,
+      "step": 54840
+    },
+    {
+      "epoch": 164.71,
+      "grad_norm": 20.582605361938477,
+      "learning_rate": 4.50950950950951e-06,
+      "loss": 0.4061,
+      "step": 54850
+    },
+    {
+      "epoch": 164.74,
+      "grad_norm": 17.31618309020996,
+      "learning_rate": 4.508508508508509e-06,
+      "loss": 0.3542,
+      "step": 54860
+    },
+    {
+      "epoch": 164.77,
+      "grad_norm": 15.660527229309082,
+      "learning_rate": 4.507507507507508e-06,
+      "loss": 0.3903,
+      "step": 54870
+    },
+    {
+      "epoch": 164.8,
+      "grad_norm": 16.76749610900879,
+      "learning_rate": 4.506506506506507e-06,
+      "loss": 0.383,
+      "step": 54880
+    },
+    {
+      "epoch": 164.83,
+      "grad_norm": 16.879011154174805,
+      "learning_rate": 4.505505505505506e-06,
+      "loss": 0.433,
+      "step": 54890
+    },
+    {
+      "epoch": 164.86,
+      "grad_norm": 14.385297775268555,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 0.3949,
+      "step": 54900
+    },
+    {
+      "epoch": 164.89,
+      "grad_norm": 13.191610336303711,
+      "learning_rate": 4.503503503503504e-06,
+      "loss": 0.4121,
+      "step": 54910
+    },
+    {
+      "epoch": 164.92,
+      "grad_norm": 20.00244140625,
+      "learning_rate": 4.502502502502503e-06,
+      "loss": 0.4038,
+      "step": 54920
+    },
+    {
+      "epoch": 164.95,
+      "grad_norm": 18.63714599609375,
+      "learning_rate": 4.501501501501501e-06,
+      "loss": 0.4054,
+      "step": 54930
+    },
+    {
+      "epoch": 164.98,
+      "grad_norm": 18.634294509887695,
+      "learning_rate": 4.500500500500501e-06,
+      "loss": 0.3829,
+      "step": 54940
+    },
+    {
+      "epoch": 165.0,
+      "eval_accuracy": 0.8652,
+      "eval_loss": 0.5282136797904968,
+      "eval_runtime": 12.8573,
+      "eval_samples_per_second": 777.769,
+      "eval_steps_per_second": 3.111,
+      "step": 54945
+    },
+    {
+      "epoch": 165.02,
+      "grad_norm": 19.605371475219727,
+      "learning_rate": 4.4994994994995e-06,
+      "loss": 0.3109,
+      "step": 54950
+    },
+    {
+      "epoch": 165.05,
+      "grad_norm": 16.851287841796875,
+      "learning_rate": 4.498498498498499e-06,
+      "loss": 0.3991,
+      "step": 54960
+    },
+    {
+      "epoch": 165.08,
+      "grad_norm": 19.71284294128418,
+      "learning_rate": 4.4974974974974974e-06,
+      "loss": 0.4147,
+      "step": 54970
+    },
+    {
+      "epoch": 165.11,
+      "grad_norm": 18.995141983032227,
+      "learning_rate": 4.496496496496497e-06,
+      "loss": 0.453,
+      "step": 54980
+    },
+    {
+      "epoch": 165.14,
+      "grad_norm": 17.508895874023438,
+      "learning_rate": 4.495495495495496e-06,
+      "loss": 0.4204,
+      "step": 54990
+    },
+    {
+      "epoch": 165.17,
+      "grad_norm": 12.46998405456543,
+      "learning_rate": 4.494494494494495e-06,
+      "loss": 0.3709,
+      "step": 55000
+    },
+    {
+      "epoch": 165.2,
+      "grad_norm": 10.211468696594238,
+      "learning_rate": 4.4934934934934935e-06,
+      "loss": 0.4286,
+      "step": 55010
+    },
+    {
+      "epoch": 165.23,
+      "grad_norm": 13.976837158203125,
+      "learning_rate": 4.492492492492493e-06,
+      "loss": 0.3801,
+      "step": 55020
+    },
+    {
+      "epoch": 165.26,
+      "grad_norm": 17.69220542907715,
+      "learning_rate": 4.491491491491492e-06,
+      "loss": 0.3844,
+      "step": 55030
+    },
+    {
+      "epoch": 165.29,
+      "grad_norm": 15.880172729492188,
+      "learning_rate": 4.490490490490491e-06,
+      "loss": 0.4117,
+      "step": 55040
+    },
+    {
+      "epoch": 165.32,
+      "grad_norm": 20.679534912109375,
+      "learning_rate": 4.48948948948949e-06,
+      "loss": 0.4022,
+      "step": 55050
+    },
+    {
+      "epoch": 165.35,
+      "grad_norm": 16.717510223388672,
+      "learning_rate": 4.488488488488489e-06,
+      "loss": 0.4286,
+      "step": 55060
+    },
+    {
+      "epoch": 165.38,
+      "grad_norm": 15.348010063171387,
+      "learning_rate": 4.487487487487488e-06,
+      "loss": 0.3275,
+      "step": 55070
+    },
+    {
+      "epoch": 165.41,
+      "grad_norm": 18.979677200317383,
+      "learning_rate": 4.486486486486487e-06,
+      "loss": 0.3814,
+      "step": 55080
+    },
+    {
+      "epoch": 165.44,
+      "grad_norm": 12.713050842285156,
+      "learning_rate": 4.485485485485486e-06,
+      "loss": 0.3462,
+      "step": 55090
+    },
+    {
+      "epoch": 165.47,
+      "grad_norm": 15.79403018951416,
+      "learning_rate": 4.484484484484485e-06,
+      "loss": 0.3864,
+      "step": 55100
+    },
+    {
+      "epoch": 165.5,
+      "grad_norm": 13.054332733154297,
+      "learning_rate": 4.483483483483484e-06,
+      "loss": 0.34,
+      "step": 55110
+    },
+    {
+      "epoch": 165.53,
+      "grad_norm": 19.75365447998047,
+      "learning_rate": 4.482482482482483e-06,
+      "loss": 0.4069,
+      "step": 55120
+    },
+    {
+      "epoch": 165.56,
+      "grad_norm": 17.507076263427734,
+      "learning_rate": 4.481481481481482e-06,
+      "loss": 0.3435,
+      "step": 55130
+    },
+    {
+      "epoch": 165.59,
+      "grad_norm": 14.217220306396484,
+      "learning_rate": 4.480480480480481e-06,
+      "loss": 0.4045,
+      "step": 55140
+    },
+    {
+      "epoch": 165.62,
+      "grad_norm": 17.861003875732422,
+      "learning_rate": 4.47947947947948e-06,
+      "loss": 0.3604,
+      "step": 55150
+    },
+    {
+      "epoch": 165.65,
+      "grad_norm": 13.404272079467773,
+      "learning_rate": 4.478478478478479e-06,
+      "loss": 0.3786,
+      "step": 55160
+    },
+    {
+      "epoch": 165.68,
+      "grad_norm": 14.263978958129883,
+      "learning_rate": 4.477477477477478e-06,
+      "loss": 0.3617,
+      "step": 55170
+    },
+    {
+      "epoch": 165.71,
+      "grad_norm": 10.393213272094727,
+      "learning_rate": 4.476476476476476e-06,
+      "loss": 0.3423,
+      "step": 55180
+    },
+    {
+      "epoch": 165.74,
+      "grad_norm": 18.60297393798828,
+      "learning_rate": 4.475475475475476e-06,
+      "loss": 0.3297,
+      "step": 55190
+    },
+    {
+      "epoch": 165.77,
+      "grad_norm": 22.285146713256836,
+      "learning_rate": 4.474474474474475e-06,
+      "loss": 0.3819,
+      "step": 55200
+    },
+    {
+      "epoch": 165.8,
+      "grad_norm": 15.247559547424316,
+      "learning_rate": 4.473473473473474e-06,
+      "loss": 0.3849,
+      "step": 55210
+    },
+    {
+      "epoch": 165.83,
+      "grad_norm": 16.095849990844727,
+      "learning_rate": 4.4724724724724725e-06,
+      "loss": 0.3482,
+      "step": 55220
+    },
+    {
+      "epoch": 165.86,
+      "grad_norm": 12.286494255065918,
+      "learning_rate": 4.471471471471472e-06,
+      "loss": 0.4264,
+      "step": 55230
+    },
+    {
+      "epoch": 165.89,
+      "grad_norm": 22.655908584594727,
+      "learning_rate": 4.470470470470471e-06,
+      "loss": 0.3227,
+      "step": 55240
+    },
+    {
+      "epoch": 165.92,
+      "grad_norm": 15.770001411437988,
+      "learning_rate": 4.46946946946947e-06,
+      "loss": 0.3604,
+      "step": 55250
+    },
+    {
+      "epoch": 165.95,
+      "grad_norm": 16.25628662109375,
+      "learning_rate": 4.4684684684684686e-06,
+      "loss": 0.4125,
+      "step": 55260
+    },
+    {
+      "epoch": 165.98,
+      "grad_norm": 12.592716217041016,
+      "learning_rate": 4.467467467467468e-06,
+      "loss": 0.3415,
+      "step": 55270
+    },
+    {
+      "epoch": 166.0,
+      "eval_accuracy": 0.8654,
+      "eval_loss": 0.5248965620994568,
+      "eval_runtime": 13.0609,
+      "eval_samples_per_second": 765.644,
+      "eval_steps_per_second": 3.063,
+      "step": 55278
+    },
+    {
+      "epoch": 166.01,
+      "grad_norm": 12.73310661315918,
+      "learning_rate": 4.466466466466467e-06,
+      "loss": 0.3277,
+      "step": 55280
+    },
+    {
+      "epoch": 166.04,
+      "grad_norm": 10.100533485412598,
+      "learning_rate": 4.465465465465465e-06,
+      "loss": 0.3784,
+      "step": 55290
+    },
+    {
+      "epoch": 166.07,
+      "grad_norm": 14.607674598693848,
+      "learning_rate": 4.464464464464465e-06,
+      "loss": 0.4066,
+      "step": 55300
+    },
+    {
+      "epoch": 166.1,
+      "grad_norm": 16.76295280456543,
+      "learning_rate": 4.463463463463464e-06,
+      "loss": 0.3817,
+      "step": 55310
+    },
+    {
+      "epoch": 166.13,
+      "grad_norm": 16.393484115600586,
+      "learning_rate": 4.462462462462463e-06,
+      "loss": 0.3759,
+      "step": 55320
+    },
+    {
+      "epoch": 166.16,
+      "grad_norm": 19.5670223236084,
+      "learning_rate": 4.461461461461462e-06,
+      "loss": 0.3777,
+      "step": 55330
+    },
+    {
+      "epoch": 166.19,
+      "grad_norm": 7.869080543518066,
+      "learning_rate": 4.460460460460461e-06,
+      "loss": 0.3284,
+      "step": 55340
+    },
+    {
+      "epoch": 166.22,
+      "grad_norm": 28.856618881225586,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.3629,
+      "step": 55350
+    },
+    {
+      "epoch": 166.25,
+      "grad_norm": 14.412503242492676,
+      "learning_rate": 4.458458458458459e-06,
+      "loss": 0.3696,
+      "step": 55360
+    },
+    {
+      "epoch": 166.28,
+      "grad_norm": 14.884553909301758,
+      "learning_rate": 4.457457457457458e-06,
+      "loss": 0.4118,
+      "step": 55370
+    },
+    {
+      "epoch": 166.31,
+      "grad_norm": 13.200601577758789,
+      "learning_rate": 4.456456456456457e-06,
+      "loss": 0.352,
+      "step": 55380
+    },
+    {
+      "epoch": 166.34,
+      "grad_norm": 19.271028518676758,
+      "learning_rate": 4.455455455455455e-06,
+      "loss": 0.4053,
+      "step": 55390
+    },
+    {
+      "epoch": 166.37,
+      "grad_norm": 14.155878067016602,
+      "learning_rate": 4.454454454454455e-06,
+      "loss": 0.4345,
+      "step": 55400
+    },
+    {
+      "epoch": 166.4,
+      "grad_norm": 17.341684341430664,
+      "learning_rate": 4.453453453453454e-06,
+      "loss": 0.3898,
+      "step": 55410
+    },
+    {
+      "epoch": 166.43,
+      "grad_norm": 19.890640258789062,
+      "learning_rate": 4.452452452452453e-06,
+      "loss": 0.3872,
+      "step": 55420
+    },
+    {
+      "epoch": 166.46,
+      "grad_norm": 15.178589820861816,
+      "learning_rate": 4.4514514514514514e-06,
+      "loss": 0.3596,
+      "step": 55430
+    },
+    {
+      "epoch": 166.49,
+      "grad_norm": 17.37016487121582,
+      "learning_rate": 4.450450450450451e-06,
+      "loss": 0.376,
+      "step": 55440
+    },
+    {
+      "epoch": 166.52,
+      "grad_norm": 16.210859298706055,
+      "learning_rate": 4.44944944944945e-06,
+      "loss": 0.3498,
+      "step": 55450
+    },
+    {
+      "epoch": 166.55,
+      "grad_norm": 13.062203407287598,
+      "learning_rate": 4.448448448448449e-06,
+      "loss": 0.3728,
+      "step": 55460
+    },
+    {
+      "epoch": 166.58,
+      "grad_norm": 16.008052825927734,
+      "learning_rate": 4.4474474474474475e-06,
+      "loss": 0.3603,
+      "step": 55470
+    },
+    {
+      "epoch": 166.61,
+      "grad_norm": 10.34046459197998,
+      "learning_rate": 4.446446446446447e-06,
+      "loss": 0.3962,
+      "step": 55480
+    },
+    {
+      "epoch": 166.64,
+      "grad_norm": 17.331249237060547,
+      "learning_rate": 4.445445445445446e-06,
+      "loss": 0.3616,
+      "step": 55490
+    },
+    {
+      "epoch": 166.67,
+      "grad_norm": 14.60738754272461,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.4227,
+      "step": 55500
+    },
+    {
+      "epoch": 166.7,
+      "grad_norm": 12.728869438171387,
+      "learning_rate": 4.443443443443444e-06,
+      "loss": 0.3503,
+      "step": 55510
+    },
+    {
+      "epoch": 166.73,
+      "grad_norm": 18.263696670532227,
+      "learning_rate": 4.442442442442443e-06,
+      "loss": 0.3971,
+      "step": 55520
+    },
+    {
+      "epoch": 166.76,
+      "grad_norm": 16.099504470825195,
+      "learning_rate": 4.441441441441442e-06,
+      "loss": 0.3911,
+      "step": 55530
+    },
+    {
+      "epoch": 166.79,
+      "grad_norm": 12.058635711669922,
+      "learning_rate": 4.44044044044044e-06,
+      "loss": 0.3536,
+      "step": 55540
+    },
+    {
+      "epoch": 166.82,
+      "grad_norm": 13.131985664367676,
+      "learning_rate": 4.43943943943944e-06,
+      "loss": 0.3523,
+      "step": 55550
+    },
+    {
+      "epoch": 166.85,
+      "grad_norm": 12.069098472595215,
+      "learning_rate": 4.438438438438439e-06,
+      "loss": 0.3526,
+      "step": 55560
+    },
+    {
+      "epoch": 166.88,
+      "grad_norm": 12.440069198608398,
+      "learning_rate": 4.437437437437438e-06,
+      "loss": 0.3495,
+      "step": 55570
+    },
+    {
+      "epoch": 166.91,
+      "grad_norm": 13.529293060302734,
+      "learning_rate": 4.436436436436437e-06,
+      "loss": 0.4065,
+      "step": 55580
+    },
+    {
+      "epoch": 166.94,
+      "grad_norm": 16.353038787841797,
+      "learning_rate": 4.435435435435436e-06,
+      "loss": 0.4199,
+      "step": 55590
+    },
+    {
+      "epoch": 166.97,
+      "grad_norm": 21.444868087768555,
+      "learning_rate": 4.434434434434435e-06,
+      "loss": 0.3636,
+      "step": 55600
+    },
+    {
+      "epoch": 167.0,
+      "grad_norm": 19.896556854248047,
+      "learning_rate": 4.433433433433434e-06,
+      "loss": 0.3599,
+      "step": 55610
+    },
+    {
+      "epoch": 167.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5251697301864624,
+      "eval_runtime": 12.8239,
+      "eval_samples_per_second": 779.795,
+      "eval_steps_per_second": 3.119,
+      "step": 55611
+    },
+    {
+      "epoch": 167.03,
+      "grad_norm": 15.800450325012207,
+      "learning_rate": 4.432432432432433e-06,
+      "loss": 0.3416,
+      "step": 55620
+    },
+    {
+      "epoch": 167.06,
+      "grad_norm": 14.296241760253906,
+      "learning_rate": 4.431431431431432e-06,
+      "loss": 0.3962,
+      "step": 55630
+    },
+    {
+      "epoch": 167.09,
+      "grad_norm": 18.51743507385254,
+      "learning_rate": 4.43043043043043e-06,
+      "loss": 0.3596,
+      "step": 55640
+    },
+    {
+      "epoch": 167.12,
+      "grad_norm": 12.96612548828125,
+      "learning_rate": 4.42942942942943e-06,
+      "loss": 0.3844,
+      "step": 55650
+    },
+    {
+      "epoch": 167.15,
+      "grad_norm": 32.8157958984375,
+      "learning_rate": 4.428428428428429e-06,
+      "loss": 0.3284,
+      "step": 55660
+    },
+    {
+      "epoch": 167.18,
+      "grad_norm": 13.51891803741455,
+      "learning_rate": 4.427427427427428e-06,
+      "loss": 0.4526,
+      "step": 55670
+    },
+    {
+      "epoch": 167.21,
+      "grad_norm": 14.83580207824707,
+      "learning_rate": 4.4264264264264265e-06,
+      "loss": 0.3447,
+      "step": 55680
+    },
+    {
+      "epoch": 167.24,
+      "grad_norm": 16.207813262939453,
+      "learning_rate": 4.425425425425426e-06,
+      "loss": 0.3355,
+      "step": 55690
+    },
+    {
+      "epoch": 167.27,
+      "grad_norm": 15.332944869995117,
+      "learning_rate": 4.424424424424425e-06,
+      "loss": 0.3894,
+      "step": 55700
+    },
+    {
+      "epoch": 167.3,
+      "grad_norm": 13.20614242553711,
+      "learning_rate": 4.423423423423424e-06,
+      "loss": 0.3519,
+      "step": 55710
+    },
+    {
+      "epoch": 167.33,
+      "grad_norm": 17.022315979003906,
+      "learning_rate": 4.4224224224224226e-06,
+      "loss": 0.3715,
+      "step": 55720
+    },
+    {
+      "epoch": 167.36,
+      "grad_norm": 17.156259536743164,
+      "learning_rate": 4.421421421421422e-06,
+      "loss": 0.4341,
+      "step": 55730
+    },
+    {
+      "epoch": 167.39,
+      "grad_norm": 18.42893409729004,
+      "learning_rate": 4.420420420420421e-06,
+      "loss": 0.4244,
+      "step": 55740
+    },
+    {
+      "epoch": 167.42,
+      "grad_norm": 12.126134872436523,
+      "learning_rate": 4.419419419419419e-06,
+      "loss": 0.3574,
+      "step": 55750
+    },
+    {
+      "epoch": 167.45,
+      "grad_norm": 10.97034740447998,
+      "learning_rate": 4.418418418418419e-06,
+      "loss": 0.369,
+      "step": 55760
+    },
+    {
+      "epoch": 167.48,
+      "grad_norm": 21.429656982421875,
+      "learning_rate": 4.417417417417418e-06,
+      "loss": 0.3838,
+      "step": 55770
+    },
+    {
+      "epoch": 167.51,
+      "grad_norm": 19.854475021362305,
+      "learning_rate": 4.416416416416417e-06,
+      "loss": 0.3645,
+      "step": 55780
+    },
+    {
+      "epoch": 167.54,
+      "grad_norm": 13.383658409118652,
+      "learning_rate": 4.415415415415415e-06,
+      "loss": 0.3554,
+      "step": 55790
+    },
+    {
+      "epoch": 167.57,
+      "grad_norm": 20.512693405151367,
+      "learning_rate": 4.414414414414415e-06,
+      "loss": 0.3746,
+      "step": 55800
+    },
+    {
+      "epoch": 167.6,
+      "grad_norm": 11.64169692993164,
+      "learning_rate": 4.413413413413414e-06,
+      "loss": 0.3881,
+      "step": 55810
+    },
+    {
+      "epoch": 167.63,
+      "grad_norm": 13.217456817626953,
+      "learning_rate": 4.412412412412413e-06,
+      "loss": 0.3713,
+      "step": 55820
+    },
+    {
+      "epoch": 167.66,
+      "grad_norm": 21.65718650817871,
+      "learning_rate": 4.411411411411412e-06,
+      "loss": 0.4479,
+      "step": 55830
+    },
+    {
+      "epoch": 167.69,
+      "grad_norm": 13.174099922180176,
+      "learning_rate": 4.410410410410411e-06,
+      "loss": 0.3484,
+      "step": 55840
+    },
+    {
+      "epoch": 167.72,
+      "grad_norm": 28.72283363342285,
+      "learning_rate": 4.409409409409409e-06,
+      "loss": 0.3462,
+      "step": 55850
+    },
+    {
+      "epoch": 167.75,
+      "grad_norm": 14.542431831359863,
+      "learning_rate": 4.408408408408409e-06,
+      "loss": 0.3969,
+      "step": 55860
+    },
+    {
+      "epoch": 167.78,
+      "grad_norm": 21.377288818359375,
+      "learning_rate": 4.407407407407408e-06,
+      "loss": 0.4192,
+      "step": 55870
+    },
+    {
+      "epoch": 167.81,
+      "grad_norm": 8.494498252868652,
+      "learning_rate": 4.406406406406407e-06,
+      "loss": 0.3943,
+      "step": 55880
+    },
+    {
+      "epoch": 167.84,
+      "grad_norm": 15.957542419433594,
+      "learning_rate": 4.4054054054054054e-06,
+      "loss": 0.3696,
+      "step": 55890
+    },
+    {
+      "epoch": 167.87,
+      "grad_norm": 14.825321197509766,
+      "learning_rate": 4.404404404404405e-06,
+      "loss": 0.4059,
+      "step": 55900
+    },
+    {
+      "epoch": 167.9,
+      "grad_norm": 11.927699089050293,
+      "learning_rate": 4.403403403403404e-06,
+      "loss": 0.4404,
+      "step": 55910
+    },
+    {
+      "epoch": 167.93,
+      "grad_norm": 13.208398818969727,
+      "learning_rate": 4.402402402402403e-06,
+      "loss": 0.3899,
+      "step": 55920
+    },
+    {
+      "epoch": 167.96,
+      "grad_norm": 15.276449203491211,
+      "learning_rate": 4.4014014014014015e-06,
+      "loss": 0.3887,
+      "step": 55930
+    },
+    {
+      "epoch": 167.99,
+      "grad_norm": 16.646677017211914,
+      "learning_rate": 4.400400400400401e-06,
+      "loss": 0.3705,
+      "step": 55940
+    },
+    {
+      "epoch": 168.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5300699472427368,
+      "eval_runtime": 12.7802,
+      "eval_samples_per_second": 782.46,
+      "eval_steps_per_second": 3.13,
+      "step": 55944
+    },
+    {
+      "epoch": 168.02,
+      "grad_norm": 22.01471519470215,
+      "learning_rate": 4.3993993993993996e-06,
+      "loss": 0.3755,
+      "step": 55950
+    },
+    {
+      "epoch": 168.05,
+      "grad_norm": 25.14999008178711,
+      "learning_rate": 4.398398398398398e-06,
+      "loss": 0.3907,
+      "step": 55960
+    },
+    {
+      "epoch": 168.08,
+      "grad_norm": 13.732797622680664,
+      "learning_rate": 4.397397397397398e-06,
+      "loss": 0.4065,
+      "step": 55970
+    },
+    {
+      "epoch": 168.11,
+      "grad_norm": 19.152921676635742,
+      "learning_rate": 4.396396396396397e-06,
+      "loss": 0.3834,
+      "step": 55980
+    },
+    {
+      "epoch": 168.14,
+      "grad_norm": 26.38957977294922,
+      "learning_rate": 4.395395395395396e-06,
+      "loss": 0.3962,
+      "step": 55990
+    },
+    {
+      "epoch": 168.17,
+      "grad_norm": 20.075016021728516,
+      "learning_rate": 4.394394394394394e-06,
+      "loss": 0.3651,
+      "step": 56000
+    },
+    {
+      "epoch": 168.2,
+      "grad_norm": 15.842062950134277,
+      "learning_rate": 4.393393393393394e-06,
+      "loss": 0.3582,
+      "step": 56010
+    },
+    {
+      "epoch": 168.23,
+      "grad_norm": 13.212736129760742,
+      "learning_rate": 4.392392392392393e-06,
+      "loss": 0.4859,
+      "step": 56020
+    },
+    {
+      "epoch": 168.26,
+      "grad_norm": 18.98668670654297,
+      "learning_rate": 4.391391391391392e-06,
+      "loss": 0.3655,
+      "step": 56030
+    },
+    {
+      "epoch": 168.29,
+      "grad_norm": 12.46191692352295,
+      "learning_rate": 4.39039039039039e-06,
+      "loss": 0.4225,
+      "step": 56040
+    },
+    {
+      "epoch": 168.32,
+      "grad_norm": 17.103851318359375,
+      "learning_rate": 4.38938938938939e-06,
+      "loss": 0.3609,
+      "step": 56050
+    },
+    {
+      "epoch": 168.35,
+      "grad_norm": 18.102779388427734,
+      "learning_rate": 4.388388388388389e-06,
+      "loss": 0.3752,
+      "step": 56060
+    },
+    {
+      "epoch": 168.38,
+      "grad_norm": 15.273604393005371,
+      "learning_rate": 4.387387387387388e-06,
+      "loss": 0.3625,
+      "step": 56070
+    },
+    {
+      "epoch": 168.41,
+      "grad_norm": 16.50876235961914,
+      "learning_rate": 4.386386386386386e-06,
+      "loss": 0.4074,
+      "step": 56080
+    },
+    {
+      "epoch": 168.44,
+      "grad_norm": 12.794107437133789,
+      "learning_rate": 4.385385385385386e-06,
+      "loss": 0.3418,
+      "step": 56090
+    },
+    {
+      "epoch": 168.47,
+      "grad_norm": 18.9176082611084,
+      "learning_rate": 4.384384384384384e-06,
+      "loss": 0.3955,
+      "step": 56100
+    },
+    {
+      "epoch": 168.5,
+      "grad_norm": 16.422706604003906,
+      "learning_rate": 4.383383383383384e-06,
+      "loss": 0.4136,
+      "step": 56110
+    },
+    {
+      "epoch": 168.53,
+      "grad_norm": 8.396899223327637,
+      "learning_rate": 4.382382382382383e-06,
+      "loss": 0.3407,
+      "step": 56120
+    },
+    {
+      "epoch": 168.56,
+      "grad_norm": 30.496013641357422,
+      "learning_rate": 4.381381381381382e-06,
+      "loss": 0.4101,
+      "step": 56130
+    },
+    {
+      "epoch": 168.59,
+      "grad_norm": 15.89106273651123,
+      "learning_rate": 4.3803803803803805e-06,
+      "loss": 0.3873,
+      "step": 56140
+    },
+    {
+      "epoch": 168.62,
+      "grad_norm": 18.845901489257812,
+      "learning_rate": 4.37937937937938e-06,
+      "loss": 0.3886,
+      "step": 56150
+    },
+    {
+      "epoch": 168.65,
+      "grad_norm": 17.462129592895508,
+      "learning_rate": 4.378378378378379e-06,
+      "loss": 0.3447,
+      "step": 56160
+    },
+    {
+      "epoch": 168.68,
+      "grad_norm": 12.9675931930542,
+      "learning_rate": 4.377377377377378e-06,
+      "loss": 0.4155,
+      "step": 56170
+    },
+    {
+      "epoch": 168.71,
+      "grad_norm": 19.45269012451172,
+      "learning_rate": 4.3763763763763765e-06,
+      "loss": 0.3324,
+      "step": 56180
+    },
+    {
+      "epoch": 168.74,
+      "grad_norm": 10.469100952148438,
+      "learning_rate": 4.375375375375376e-06,
+      "loss": 0.3553,
+      "step": 56190
+    },
+    {
+      "epoch": 168.77,
+      "grad_norm": 15.421512603759766,
+      "learning_rate": 4.374374374374375e-06,
+      "loss": 0.355,
+      "step": 56200
+    },
+    {
+      "epoch": 168.8,
+      "grad_norm": 21.223556518554688,
+      "learning_rate": 4.373373373373373e-06,
+      "loss": 0.3551,
+      "step": 56210
+    },
+    {
+      "epoch": 168.83,
+      "grad_norm": 12.923812866210938,
+      "learning_rate": 4.372372372372373e-06,
+      "loss": 0.3806,
+      "step": 56220
+    },
+    {
+      "epoch": 168.86,
+      "grad_norm": 17.667848587036133,
+      "learning_rate": 4.371371371371372e-06,
+      "loss": 0.3536,
+      "step": 56230
+    },
+    {
+      "epoch": 168.89,
+      "grad_norm": 13.803996086120605,
+      "learning_rate": 4.370370370370371e-06,
+      "loss": 0.3735,
+      "step": 56240
+    },
+    {
+      "epoch": 168.92,
+      "grad_norm": 13.809103965759277,
+      "learning_rate": 4.369369369369369e-06,
+      "loss": 0.4001,
+      "step": 56250
+    },
+    {
+      "epoch": 168.95,
+      "grad_norm": 17.825586318969727,
+      "learning_rate": 4.368368368368369e-06,
+      "loss": 0.3719,
+      "step": 56260
+    },
+    {
+      "epoch": 168.98,
+      "grad_norm": 17.908336639404297,
+      "learning_rate": 4.367367367367368e-06,
+      "loss": 0.4122,
+      "step": 56270
+    },
+    {
+      "epoch": 169.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.5357725620269775,
+      "eval_runtime": 12.7411,
+      "eval_samples_per_second": 784.863,
+      "eval_steps_per_second": 3.139,
+      "step": 56277
+    },
+    {
+      "epoch": 169.01,
+      "grad_norm": 12.425196647644043,
+      "learning_rate": 4.366366366366367e-06,
+      "loss": 0.5141,
+      "step": 56280
+    },
+    {
+      "epoch": 169.04,
+      "grad_norm": 10.186840057373047,
+      "learning_rate": 4.365365365365365e-06,
+      "loss": 0.3242,
+      "step": 56290
+    },
+    {
+      "epoch": 169.07,
+      "grad_norm": 14.42274284362793,
+      "learning_rate": 4.364364364364365e-06,
+      "loss": 0.4264,
+      "step": 56300
+    },
+    {
+      "epoch": 169.1,
+      "grad_norm": 12.531476020812988,
+      "learning_rate": 4.363363363363363e-06,
+      "loss": 0.4397,
+      "step": 56310
+    },
+    {
+      "epoch": 169.13,
+      "grad_norm": 13.007606506347656,
+      "learning_rate": 4.362362362362363e-06,
+      "loss": 0.3968,
+      "step": 56320
+    },
+    {
+      "epoch": 169.16,
+      "grad_norm": 13.926525115966797,
+      "learning_rate": 4.361361361361361e-06,
+      "loss": 0.3585,
+      "step": 56330
+    },
+    {
+      "epoch": 169.19,
+      "grad_norm": 12.366735458374023,
+      "learning_rate": 4.360360360360361e-06,
+      "loss": 0.407,
+      "step": 56340
+    },
+    {
+      "epoch": 169.22,
+      "grad_norm": 15.59043025970459,
+      "learning_rate": 4.3593593593593594e-06,
+      "loss": 0.3202,
+      "step": 56350
+    },
+    {
+      "epoch": 169.25,
+      "grad_norm": 15.17165756225586,
+      "learning_rate": 4.358358358358359e-06,
+      "loss": 0.3241,
+      "step": 56360
+    },
+    {
+      "epoch": 169.28,
+      "grad_norm": 18.283876419067383,
+      "learning_rate": 4.357357357357358e-06,
+      "loss": 0.3976,
+      "step": 56370
+    },
+    {
+      "epoch": 169.31,
+      "grad_norm": 14.6078519821167,
+      "learning_rate": 4.356356356356357e-06,
+      "loss": 0.3904,
+      "step": 56380
+    },
+    {
+      "epoch": 169.34,
+      "grad_norm": 11.528279304504395,
+      "learning_rate": 4.3553553553553555e-06,
+      "loss": 0.382,
+      "step": 56390
+    },
+    {
+      "epoch": 169.37,
+      "grad_norm": 15.306736946105957,
+      "learning_rate": 4.354354354354355e-06,
+      "loss": 0.4439,
+      "step": 56400
+    },
+    {
+      "epoch": 169.4,
+      "grad_norm": 15.499312400817871,
+      "learning_rate": 4.3533533533533535e-06,
+      "loss": 0.3835,
+      "step": 56410
+    },
+    {
+      "epoch": 169.43,
+      "grad_norm": 15.241299629211426,
+      "learning_rate": 4.352352352352352e-06,
+      "loss": 0.3901,
+      "step": 56420
+    },
+    {
+      "epoch": 169.46,
+      "grad_norm": 17.96285057067871,
+      "learning_rate": 4.351351351351352e-06,
+      "loss": 0.3879,
+      "step": 56430
+    },
+    {
+      "epoch": 169.49,
+      "grad_norm": 14.373953819274902,
+      "learning_rate": 4.350350350350351e-06,
+      "loss": 0.4139,
+      "step": 56440
+    },
+    {
+      "epoch": 169.52,
+      "grad_norm": 25.730083465576172,
+      "learning_rate": 4.34934934934935e-06,
+      "loss": 0.3727,
+      "step": 56450
+    },
+    {
+      "epoch": 169.55,
+      "grad_norm": 20.238168716430664,
+      "learning_rate": 4.348348348348348e-06,
+      "loss": 0.4207,
+      "step": 56460
+    },
+    {
+      "epoch": 169.58,
+      "grad_norm": 14.326674461364746,
+      "learning_rate": 4.347347347347348e-06,
+      "loss": 0.4165,
+      "step": 56470
+    },
+    {
+      "epoch": 169.61,
+      "grad_norm": 13.78958797454834,
+      "learning_rate": 4.346346346346347e-06,
+      "loss": 0.452,
+      "step": 56480
+    },
+    {
+      "epoch": 169.64,
+      "grad_norm": 21.098142623901367,
+      "learning_rate": 4.345345345345346e-06,
+      "loss": 0.3358,
+      "step": 56490
+    },
+    {
+      "epoch": 169.67,
+      "grad_norm": 15.661078453063965,
+      "learning_rate": 4.344344344344344e-06,
+      "loss": 0.3696,
+      "step": 56500
+    },
+    {
+      "epoch": 169.7,
+      "grad_norm": 11.455445289611816,
+      "learning_rate": 4.343343343343344e-06,
+      "loss": 0.3426,
+      "step": 56510
+    },
+    {
+      "epoch": 169.73,
+      "grad_norm": 14.869673728942871,
+      "learning_rate": 4.342342342342343e-06,
+      "loss": 0.3928,
+      "step": 56520
+    },
+    {
+      "epoch": 169.76,
+      "grad_norm": 15.880226135253906,
+      "learning_rate": 4.341341341341342e-06,
+      "loss": 0.3784,
+      "step": 56530
+    },
+    {
+      "epoch": 169.79,
+      "grad_norm": 16.56886100769043,
+      "learning_rate": 4.34034034034034e-06,
+      "loss": 0.3882,
+      "step": 56540
+    },
+    {
+      "epoch": 169.82,
+      "grad_norm": 17.10436248779297,
+      "learning_rate": 4.33933933933934e-06,
+      "loss": 0.3978,
+      "step": 56550
+    },
+    {
+      "epoch": 169.85,
+      "grad_norm": 25.78338623046875,
+      "learning_rate": 4.338338338338338e-06,
+      "loss": 0.3629,
+      "step": 56560
+    },
+    {
+      "epoch": 169.88,
+      "grad_norm": 14.37177562713623,
+      "learning_rate": 4.337337337337338e-06,
+      "loss": 0.4097,
+      "step": 56570
+    },
+    {
+      "epoch": 169.91,
+      "grad_norm": 13.646896362304688,
+      "learning_rate": 4.3363363363363364e-06,
+      "loss": 0.3691,
+      "step": 56580
+    },
+    {
+      "epoch": 169.94,
+      "grad_norm": 22.690093994140625,
+      "learning_rate": 4.335335335335336e-06,
+      "loss": 0.3655,
+      "step": 56590
+    },
+    {
+      "epoch": 169.97,
+      "grad_norm": 19.810792922973633,
+      "learning_rate": 4.3343343343343345e-06,
+      "loss": 0.3911,
+      "step": 56600
+    },
+    {
+      "epoch": 170.0,
+      "grad_norm": 18.38820457458496,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 0.3473,
+      "step": 56610
+    },
+    {
+      "epoch": 170.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.5297730565071106,
+      "eval_runtime": 13.1524,
+      "eval_samples_per_second": 760.317,
+      "eval_steps_per_second": 3.041,
+      "step": 56610
+    },
+    {
+      "epoch": 170.03,
+      "grad_norm": 20.561038970947266,
+      "learning_rate": 4.3323323323323325e-06,
+      "loss": 0.3484,
+      "step": 56620
+    },
+    {
+      "epoch": 170.06,
+      "grad_norm": 14.18109130859375,
+      "learning_rate": 4.331331331331332e-06,
+      "loss": 0.3745,
+      "step": 56630
+    },
+    {
+      "epoch": 170.09,
+      "grad_norm": 17.069900512695312,
+      "learning_rate": 4.3303303303303305e-06,
+      "loss": 0.3855,
+      "step": 56640
+    },
+    {
+      "epoch": 170.12,
+      "grad_norm": 17.644935607910156,
+      "learning_rate": 4.32932932932933e-06,
+      "loss": 0.3405,
+      "step": 56650
+    },
+    {
+      "epoch": 170.15,
+      "grad_norm": 10.872136116027832,
+      "learning_rate": 4.328328328328329e-06,
+      "loss": 0.3558,
+      "step": 56660
+    },
+    {
+      "epoch": 170.18,
+      "grad_norm": 25.831695556640625,
+      "learning_rate": 4.327327327327327e-06,
+      "loss": 0.4106,
+      "step": 56670
+    },
+    {
+      "epoch": 170.21,
+      "grad_norm": 10.665085792541504,
+      "learning_rate": 4.326326326326327e-06,
+      "loss": 0.3311,
+      "step": 56680
+    },
+    {
+      "epoch": 170.24,
+      "grad_norm": 18.547231674194336,
+      "learning_rate": 4.325325325325326e-06,
+      "loss": 0.3794,
+      "step": 56690
+    },
+    {
+      "epoch": 170.27,
+      "grad_norm": 14.098223686218262,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.3995,
+      "step": 56700
+    },
+    {
+      "epoch": 170.3,
+      "grad_norm": 17.59161376953125,
+      "learning_rate": 4.323323323323323e-06,
+      "loss": 0.3485,
+      "step": 56710
+    },
+    {
+      "epoch": 170.33,
+      "grad_norm": 13.916496276855469,
+      "learning_rate": 4.322322322322323e-06,
+      "loss": 0.4002,
+      "step": 56720
+    },
+    {
+      "epoch": 170.36,
+      "grad_norm": 24.46206283569336,
+      "learning_rate": 4.321321321321322e-06,
+      "loss": 0.3849,
+      "step": 56730
+    },
+    {
+      "epoch": 170.39,
+      "grad_norm": 20.157501220703125,
+      "learning_rate": 4.320320320320321e-06,
+      "loss": 0.3815,
+      "step": 56740
+    },
+    {
+      "epoch": 170.42,
+      "grad_norm": 15.674138069152832,
+      "learning_rate": 4.319319319319319e-06,
+      "loss": 0.4087,
+      "step": 56750
+    },
+    {
+      "epoch": 170.45,
+      "grad_norm": 19.95004653930664,
+      "learning_rate": 4.318318318318319e-06,
+      "loss": 0.3739,
+      "step": 56760
+    },
+    {
+      "epoch": 170.48,
+      "grad_norm": 17.08576011657715,
+      "learning_rate": 4.317317317317317e-06,
+      "loss": 0.3577,
+      "step": 56770
+    },
+    {
+      "epoch": 170.51,
+      "grad_norm": 21.85826301574707,
+      "learning_rate": 4.316316316316317e-06,
+      "loss": 0.4234,
+      "step": 56780
+    },
+    {
+      "epoch": 170.54,
+      "grad_norm": 18.06846046447754,
+      "learning_rate": 4.315315315315315e-06,
+      "loss": 0.3835,
+      "step": 56790
+    },
+    {
+      "epoch": 170.57,
+      "grad_norm": 15.876794815063477,
+      "learning_rate": 4.314314314314315e-06,
+      "loss": 0.3842,
+      "step": 56800
+    },
+    {
+      "epoch": 170.6,
+      "grad_norm": 12.541007041931152,
+      "learning_rate": 4.3133133133133134e-06,
+      "loss": 0.3712,
+      "step": 56810
+    },
+    {
+      "epoch": 170.63,
+      "grad_norm": 18.40793228149414,
+      "learning_rate": 4.312312312312313e-06,
+      "loss": 0.4383,
+      "step": 56820
+    },
+    {
+      "epoch": 170.66,
+      "grad_norm": 19.352867126464844,
+      "learning_rate": 4.3113113113113115e-06,
+      "loss": 0.362,
+      "step": 56830
+    },
+    {
+      "epoch": 170.69,
+      "grad_norm": 12.005217552185059,
+      "learning_rate": 4.310310310310311e-06,
+      "loss": 0.4199,
+      "step": 56840
+    },
+    {
+      "epoch": 170.72,
+      "grad_norm": 13.93285083770752,
+      "learning_rate": 4.3093093093093095e-06,
+      "loss": 0.365,
+      "step": 56850
+    },
+    {
+      "epoch": 170.75,
+      "grad_norm": 13.212437629699707,
+      "learning_rate": 4.308308308308309e-06,
+      "loss": 0.374,
+      "step": 56860
+    },
+    {
+      "epoch": 170.78,
+      "grad_norm": 24.983875274658203,
+      "learning_rate": 4.3073073073073075e-06,
+      "loss": 0.382,
+      "step": 56870
+    },
+    {
+      "epoch": 170.81,
+      "grad_norm": 18.697402954101562,
+      "learning_rate": 4.306306306306306e-06,
+      "loss": 0.3805,
+      "step": 56880
+    },
+    {
+      "epoch": 170.84,
+      "grad_norm": 17.745403289794922,
+      "learning_rate": 4.305305305305306e-06,
+      "loss": 0.3883,
+      "step": 56890
+    },
+    {
+      "epoch": 170.87,
+      "grad_norm": 16.216272354125977,
+      "learning_rate": 4.304304304304305e-06,
+      "loss": 0.3724,
+      "step": 56900
+    },
+    {
+      "epoch": 170.9,
+      "grad_norm": 13.480070114135742,
+      "learning_rate": 4.303303303303304e-06,
+      "loss": 0.394,
+      "step": 56910
+    },
+    {
+      "epoch": 170.93,
+      "grad_norm": 19.61466407775879,
+      "learning_rate": 4.302302302302302e-06,
+      "loss": 0.3673,
+      "step": 56920
+    },
+    {
+      "epoch": 170.96,
+      "grad_norm": 25.419227600097656,
+      "learning_rate": 4.301301301301302e-06,
+      "loss": 0.4037,
+      "step": 56930
+    },
+    {
+      "epoch": 170.99,
+      "grad_norm": 20.688030242919922,
+      "learning_rate": 4.300300300300301e-06,
+      "loss": 0.3825,
+      "step": 56940
+    },
+    {
+      "epoch": 171.0,
+      "eval_accuracy": 0.8643,
+      "eval_loss": 0.5256262421607971,
+      "eval_runtime": 12.6298,
+      "eval_samples_per_second": 791.775,
+      "eval_steps_per_second": 3.167,
+      "step": 56943
+    },
+    {
+      "epoch": 171.02,
+      "grad_norm": 12.25744915008545,
+      "learning_rate": 4.2992992992993e-06,
+      "loss": 0.3065,
+      "step": 56950
+    },
+    {
+      "epoch": 171.05,
+      "grad_norm": 14.202835083007812,
+      "learning_rate": 4.298298298298298e-06,
+      "loss": 0.3839,
+      "step": 56960
+    },
+    {
+      "epoch": 171.08,
+      "grad_norm": 13.939863204956055,
+      "learning_rate": 4.297297297297298e-06,
+      "loss": 0.3767,
+      "step": 56970
+    },
+    {
+      "epoch": 171.11,
+      "grad_norm": 15.340904235839844,
+      "learning_rate": 4.296296296296296e-06,
+      "loss": 0.3589,
+      "step": 56980
+    },
+    {
+      "epoch": 171.14,
+      "grad_norm": 17.97730827331543,
+      "learning_rate": 4.295295295295296e-06,
+      "loss": 0.3849,
+      "step": 56990
+    },
+    {
+      "epoch": 171.17,
+      "grad_norm": 13.499581336975098,
+      "learning_rate": 4.294294294294294e-06,
+      "loss": 0.3866,
+      "step": 57000
+    },
+    {
+      "epoch": 171.2,
+      "grad_norm": 12.762290954589844,
+      "learning_rate": 4.293293293293294e-06,
+      "loss": 0.412,
+      "step": 57010
+    },
+    {
+      "epoch": 171.23,
+      "grad_norm": 14.749709129333496,
+      "learning_rate": 4.292292292292292e-06,
+      "loss": 0.3878,
+      "step": 57020
+    },
+    {
+      "epoch": 171.26,
+      "grad_norm": 12.713895797729492,
+      "learning_rate": 4.291291291291292e-06,
+      "loss": 0.4018,
+      "step": 57030
+    },
+    {
+      "epoch": 171.29,
+      "grad_norm": 14.898329734802246,
+      "learning_rate": 4.2902902902902904e-06,
+      "loss": 0.4039,
+      "step": 57040
+    },
+    {
+      "epoch": 171.32,
+      "grad_norm": 12.865190505981445,
+      "learning_rate": 4.28928928928929e-06,
+      "loss": 0.3562,
+      "step": 57050
+    },
+    {
+      "epoch": 171.35,
+      "grad_norm": 14.700349807739258,
+      "learning_rate": 4.2882882882882885e-06,
+      "loss": 0.4067,
+      "step": 57060
+    },
+    {
+      "epoch": 171.38,
+      "grad_norm": 14.766688346862793,
+      "learning_rate": 4.287287287287288e-06,
+      "loss": 0.3681,
+      "step": 57070
+    },
+    {
+      "epoch": 171.41,
+      "grad_norm": 14.219417572021484,
+      "learning_rate": 4.2862862862862865e-06,
+      "loss": 0.3503,
+      "step": 57080
+    },
+    {
+      "epoch": 171.44,
+      "grad_norm": 17.515365600585938,
+      "learning_rate": 4.285285285285286e-06,
+      "loss": 0.4015,
+      "step": 57090
+    },
+    {
+      "epoch": 171.47,
+      "grad_norm": 17.78846549987793,
+      "learning_rate": 4.2842842842842845e-06,
+      "loss": 0.3202,
+      "step": 57100
+    },
+    {
+      "epoch": 171.5,
+      "grad_norm": 10.458614349365234,
+      "learning_rate": 4.283283283283284e-06,
+      "loss": 0.3816,
+      "step": 57110
+    },
+    {
+      "epoch": 171.53,
+      "grad_norm": 16.334318161010742,
+      "learning_rate": 4.282282282282283e-06,
+      "loss": 0.3486,
+      "step": 57120
+    },
+    {
+      "epoch": 171.56,
+      "grad_norm": 29.326871871948242,
+      "learning_rate": 4.281281281281281e-06,
+      "loss": 0.4237,
+      "step": 57130
+    },
+    {
+      "epoch": 171.59,
+      "grad_norm": 16.021785736083984,
+      "learning_rate": 4.280280280280281e-06,
+      "loss": 0.3782,
+      "step": 57140
+    },
+    {
+      "epoch": 171.62,
+      "grad_norm": 14.200824737548828,
+      "learning_rate": 4.27927927927928e-06,
+      "loss": 0.3936,
+      "step": 57150
+    },
+    {
+      "epoch": 171.65,
+      "grad_norm": 11.844581604003906,
+      "learning_rate": 4.278278278278279e-06,
+      "loss": 0.385,
+      "step": 57160
+    },
+    {
+      "epoch": 171.68,
+      "grad_norm": 12.244013786315918,
+      "learning_rate": 4.277277277277277e-06,
+      "loss": 0.4009,
+      "step": 57170
+    },
+    {
+      "epoch": 171.71,
+      "grad_norm": 19.525978088378906,
+      "learning_rate": 4.276276276276277e-06,
+      "loss": 0.423,
+      "step": 57180
+    },
+    {
+      "epoch": 171.74,
+      "grad_norm": 11.8086519241333,
+      "learning_rate": 4.275275275275276e-06,
+      "loss": 0.3931,
+      "step": 57190
+    },
+    {
+      "epoch": 171.77,
+      "grad_norm": 19.843446731567383,
+      "learning_rate": 4.274274274274275e-06,
+      "loss": 0.3499,
+      "step": 57200
+    },
+    {
+      "epoch": 171.8,
+      "grad_norm": 21.28567123413086,
+      "learning_rate": 4.273273273273273e-06,
+      "loss": 0.3894,
+      "step": 57210
+    },
+    {
+      "epoch": 171.83,
+      "grad_norm": 15.554808616638184,
+      "learning_rate": 4.272272272272273e-06,
+      "loss": 0.3748,
+      "step": 57220
+    },
+    {
+      "epoch": 171.86,
+      "grad_norm": 13.282367706298828,
+      "learning_rate": 4.271271271271271e-06,
+      "loss": 0.4174,
+      "step": 57230
+    },
+    {
+      "epoch": 171.89,
+      "grad_norm": 20.956287384033203,
+      "learning_rate": 4.270270270270271e-06,
+      "loss": 0.4272,
+      "step": 57240
+    },
+    {
+      "epoch": 171.92,
+      "grad_norm": 14.130888938903809,
+      "learning_rate": 4.269269269269269e-06,
+      "loss": 0.3568,
+      "step": 57250
+    },
+    {
+      "epoch": 171.95,
+      "grad_norm": 22.304006576538086,
+      "learning_rate": 4.268268268268269e-06,
+      "loss": 0.3812,
+      "step": 57260
+    },
+    {
+      "epoch": 171.98,
+      "grad_norm": 10.409132957458496,
+      "learning_rate": 4.2672672672672674e-06,
+      "loss": 0.3841,
+      "step": 57270
+    },
+    {
+      "epoch": 172.0,
+      "eval_accuracy": 0.8668,
+      "eval_loss": 0.5228556394577026,
+      "eval_runtime": 12.9063,
+      "eval_samples_per_second": 774.814,
+      "eval_steps_per_second": 3.099,
+      "step": 57276
+    },
+    {
+      "epoch": 172.01,
+      "grad_norm": 16.141401290893555,
+      "learning_rate": 4.266266266266267e-06,
+      "loss": 0.4795,
+      "step": 57280
+    },
+    {
+      "epoch": 172.04,
+      "grad_norm": 17.48192596435547,
+      "learning_rate": 4.2652652652652655e-06,
+      "loss": 0.3995,
+      "step": 57290
+    },
+    {
+      "epoch": 172.07,
+      "grad_norm": 10.766530990600586,
+      "learning_rate": 4.264264264264265e-06,
+      "loss": 0.3865,
+      "step": 57300
+    },
+    {
+      "epoch": 172.1,
+      "grad_norm": 16.017885208129883,
+      "learning_rate": 4.2632632632632635e-06,
+      "loss": 0.4088,
+      "step": 57310
+    },
+    {
+      "epoch": 172.13,
+      "grad_norm": 13.196261405944824,
+      "learning_rate": 4.262262262262263e-06,
+      "loss": 0.3632,
+      "step": 57320
+    },
+    {
+      "epoch": 172.16,
+      "grad_norm": 15.06296157836914,
+      "learning_rate": 4.2612612612612615e-06,
+      "loss": 0.3698,
+      "step": 57330
+    },
+    {
+      "epoch": 172.19,
+      "grad_norm": 18.926654815673828,
+      "learning_rate": 4.26026026026026e-06,
+      "loss": 0.3009,
+      "step": 57340
+    },
+    {
+      "epoch": 172.22,
+      "grad_norm": 12.891047477722168,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 0.3337,
+      "step": 57350
+    },
+    {
+      "epoch": 172.25,
+      "grad_norm": 16.172075271606445,
+      "learning_rate": 4.258258258258259e-06,
+      "loss": 0.3453,
+      "step": 57360
+    },
+    {
+      "epoch": 172.28,
+      "grad_norm": 20.749652862548828,
+      "learning_rate": 4.257257257257258e-06,
+      "loss": 0.3348,
+      "step": 57370
+    },
+    {
+      "epoch": 172.31,
+      "grad_norm": 12.353781700134277,
+      "learning_rate": 4.256256256256256e-06,
+      "loss": 0.3577,
+      "step": 57380
+    },
+    {
+      "epoch": 172.34,
+      "grad_norm": 9.224239349365234,
+      "learning_rate": 4.255255255255256e-06,
+      "loss": 0.3468,
+      "step": 57390
+    },
+    {
+      "epoch": 172.37,
+      "grad_norm": 14.028410911560059,
+      "learning_rate": 4.254254254254255e-06,
+      "loss": 0.3329,
+      "step": 57400
+    },
+    {
+      "epoch": 172.4,
+      "grad_norm": 13.380922317504883,
+      "learning_rate": 4.253253253253254e-06,
+      "loss": 0.4405,
+      "step": 57410
+    },
+    {
+      "epoch": 172.43,
+      "grad_norm": 19.872684478759766,
+      "learning_rate": 4.252252252252252e-06,
+      "loss": 0.3844,
+      "step": 57420
+    },
+    {
+      "epoch": 172.46,
+      "grad_norm": 21.78046226501465,
+      "learning_rate": 4.251251251251252e-06,
+      "loss": 0.3195,
+      "step": 57430
+    },
+    {
+      "epoch": 172.49,
+      "grad_norm": 14.379535675048828,
+      "learning_rate": 4.25025025025025e-06,
+      "loss": 0.3527,
+      "step": 57440
+    },
+    {
+      "epoch": 172.52,
+      "grad_norm": 13.067408561706543,
+      "learning_rate": 4.24924924924925e-06,
+      "loss": 0.3518,
+      "step": 57450
+    },
+    {
+      "epoch": 172.55,
+      "grad_norm": 18.766061782836914,
+      "learning_rate": 4.248248248248248e-06,
+      "loss": 0.371,
+      "step": 57460
+    },
+    {
+      "epoch": 172.58,
+      "grad_norm": 17.42868995666504,
+      "learning_rate": 4.247247247247248e-06,
+      "loss": 0.3896,
+      "step": 57470
+    },
+    {
+      "epoch": 172.61,
+      "grad_norm": 22.810535430908203,
+      "learning_rate": 4.246246246246246e-06,
+      "loss": 0.375,
+      "step": 57480
+    },
+    {
+      "epoch": 172.64,
+      "grad_norm": 19.509571075439453,
+      "learning_rate": 4.245245245245246e-06,
+      "loss": 0.4283,
+      "step": 57490
+    },
+    {
+      "epoch": 172.67,
+      "grad_norm": 14.277532577514648,
+      "learning_rate": 4.2442442442442444e-06,
+      "loss": 0.4126,
+      "step": 57500
+    },
+    {
+      "epoch": 172.7,
+      "grad_norm": 8.973878860473633,
+      "learning_rate": 4.243243243243244e-06,
+      "loss": 0.3873,
+      "step": 57510
+    },
+    {
+      "epoch": 172.73,
+      "grad_norm": 25.24574089050293,
+      "learning_rate": 4.2422422422422425e-06,
+      "loss": 0.3509,
+      "step": 57520
+    },
+    {
+      "epoch": 172.76,
+      "grad_norm": 11.894126892089844,
+      "learning_rate": 4.241241241241242e-06,
+      "loss": 0.4083,
+      "step": 57530
+    },
+    {
+      "epoch": 172.79,
+      "grad_norm": 16.47452163696289,
+      "learning_rate": 4.2402402402402405e-06,
+      "loss": 0.3807,
+      "step": 57540
+    },
+    {
+      "epoch": 172.82,
+      "grad_norm": 16.612918853759766,
+      "learning_rate": 4.23923923923924e-06,
+      "loss": 0.3792,
+      "step": 57550
+    },
+    {
+      "epoch": 172.85,
+      "grad_norm": 14.219429969787598,
+      "learning_rate": 4.2382382382382385e-06,
+      "loss": 0.3576,
+      "step": 57560
+    },
+    {
+      "epoch": 172.88,
+      "grad_norm": 15.944235801696777,
+      "learning_rate": 4.237237237237238e-06,
+      "loss": 0.3643,
+      "step": 57570
+    },
+    {
+      "epoch": 172.91,
+      "grad_norm": 15.820178985595703,
+      "learning_rate": 4.2362362362362366e-06,
+      "loss": 0.339,
+      "step": 57580
+    },
+    {
+      "epoch": 172.94,
+      "grad_norm": 20.102527618408203,
+      "learning_rate": 4.235235235235235e-06,
+      "loss": 0.3737,
+      "step": 57590
+    },
+    {
+      "epoch": 172.97,
+      "grad_norm": 16.324352264404297,
+      "learning_rate": 4.234234234234235e-06,
+      "loss": 0.3543,
+      "step": 57600
+    },
+    {
+      "epoch": 173.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.5269811153411865,
+      "eval_runtime": 12.6175,
+      "eval_samples_per_second": 792.552,
+      "eval_steps_per_second": 3.17,
+      "step": 57609
+    },
+    {
+      "epoch": 173.0,
+      "grad_norm": 14.722068786621094,
+      "learning_rate": 4.233233233233234e-06,
+      "loss": 0.3683,
+      "step": 57610
+    },
+    {
+      "epoch": 173.03,
+      "grad_norm": 17.1778621673584,
+      "learning_rate": 4.232232232232233e-06,
+      "loss": 0.3656,
+      "step": 57620
+    },
+    {
+      "epoch": 173.06,
+      "grad_norm": 15.721131324768066,
+      "learning_rate": 4.231231231231231e-06,
+      "loss": 0.3569,
+      "step": 57630
+    },
+    {
+      "epoch": 173.09,
+      "grad_norm": 16.874025344848633,
+      "learning_rate": 4.230230230230231e-06,
+      "loss": 0.3516,
+      "step": 57640
+    },
+    {
+      "epoch": 173.12,
+      "grad_norm": 17.049219131469727,
+      "learning_rate": 4.22922922922923e-06,
+      "loss": 0.3954,
+      "step": 57650
+    },
+    {
+      "epoch": 173.15,
+      "grad_norm": 17.236125946044922,
+      "learning_rate": 4.228228228228229e-06,
+      "loss": 0.3467,
+      "step": 57660
+    },
+    {
+      "epoch": 173.18,
+      "grad_norm": 12.872593879699707,
+      "learning_rate": 4.227227227227227e-06,
+      "loss": 0.3115,
+      "step": 57670
+    },
+    {
+      "epoch": 173.21,
+      "grad_norm": 15.12679672241211,
+      "learning_rate": 4.226226226226227e-06,
+      "loss": 0.4214,
+      "step": 57680
+    },
+    {
+      "epoch": 173.24,
+      "grad_norm": 16.730281829833984,
+      "learning_rate": 4.225225225225225e-06,
+      "loss": 0.3474,
+      "step": 57690
+    },
+    {
+      "epoch": 173.27,
+      "grad_norm": 18.717182159423828,
+      "learning_rate": 4.224224224224225e-06,
+      "loss": 0.3515,
+      "step": 57700
+    },
+    {
+      "epoch": 173.3,
+      "grad_norm": 16.16720962524414,
+      "learning_rate": 4.223223223223223e-06,
+      "loss": 0.4081,
+      "step": 57710
+    },
+    {
+      "epoch": 173.33,
+      "grad_norm": 12.642423629760742,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 0.3597,
+      "step": 57720
+    },
+    {
+      "epoch": 173.36,
+      "grad_norm": 18.855533599853516,
+      "learning_rate": 4.2212212212212214e-06,
+      "loss": 0.3832,
+      "step": 57730
+    },
+    {
+      "epoch": 173.39,
+      "grad_norm": 26.11993408203125,
+      "learning_rate": 4.220220220220221e-06,
+      "loss": 0.3635,
+      "step": 57740
+    },
+    {
+      "epoch": 173.42,
+      "grad_norm": 15.049911499023438,
+      "learning_rate": 4.2192192192192195e-06,
+      "loss": 0.3588,
+      "step": 57750
+    },
+    {
+      "epoch": 173.45,
+      "grad_norm": 15.795404434204102,
+      "learning_rate": 4.218218218218219e-06,
+      "loss": 0.3421,
+      "step": 57760
+    },
+    {
+      "epoch": 173.48,
+      "grad_norm": 16.169479370117188,
+      "learning_rate": 4.2172172172172175e-06,
+      "loss": 0.4193,
+      "step": 57770
+    },
+    {
+      "epoch": 173.51,
+      "grad_norm": 12.967020034790039,
+      "learning_rate": 4.216216216216217e-06,
+      "loss": 0.35,
+      "step": 57780
+    },
+    {
+      "epoch": 173.54,
+      "grad_norm": 15.44735336303711,
+      "learning_rate": 4.2152152152152155e-06,
+      "loss": 0.3627,
+      "step": 57790
+    },
+    {
+      "epoch": 173.57,
+      "grad_norm": 14.469039916992188,
+      "learning_rate": 4.214214214214214e-06,
+      "loss": 0.3678,
+      "step": 57800
+    },
+    {
+      "epoch": 173.6,
+      "grad_norm": 11.469844818115234,
+      "learning_rate": 4.2132132132132136e-06,
+      "loss": 0.335,
+      "step": 57810
+    },
+    {
+      "epoch": 173.63,
+      "grad_norm": 12.479352951049805,
+      "learning_rate": 4.212212212212213e-06,
+      "loss": 0.4026,
+      "step": 57820
+    },
+    {
+      "epoch": 173.66,
+      "grad_norm": 12.23619556427002,
+      "learning_rate": 4.211211211211212e-06,
+      "loss": 0.3865,
+      "step": 57830
+    },
+    {
+      "epoch": 173.69,
+      "grad_norm": 21.697132110595703,
+      "learning_rate": 4.21021021021021e-06,
+      "loss": 0.3243,
+      "step": 57840
+    },
+    {
+      "epoch": 173.72,
+      "grad_norm": 11.666245460510254,
+      "learning_rate": 4.20920920920921e-06,
+      "loss": 0.3919,
+      "step": 57850
+    },
+    {
+      "epoch": 173.75,
+      "grad_norm": 12.43196964263916,
+      "learning_rate": 4.208208208208209e-06,
+      "loss": 0.3872,
+      "step": 57860
+    },
+    {
+      "epoch": 173.78,
+      "grad_norm": 15.85690975189209,
+      "learning_rate": 4.207207207207208e-06,
+      "loss": 0.403,
+      "step": 57870
+    },
+    {
+      "epoch": 173.81,
+      "grad_norm": 14.177299499511719,
+      "learning_rate": 4.206206206206206e-06,
+      "loss": 0.3622,
+      "step": 57880
+    },
+    {
+      "epoch": 173.84,
+      "grad_norm": 12.051791191101074,
+      "learning_rate": 4.205205205205206e-06,
+      "loss": 0.3732,
+      "step": 57890
+    },
+    {
+      "epoch": 173.87,
+      "grad_norm": 12.137227058410645,
+      "learning_rate": 4.204204204204204e-06,
+      "loss": 0.3722,
+      "step": 57900
+    },
+    {
+      "epoch": 173.9,
+      "grad_norm": 15.890396118164062,
+      "learning_rate": 4.203203203203203e-06,
+      "loss": 0.3515,
+      "step": 57910
+    },
+    {
+      "epoch": 173.93,
+      "grad_norm": 10.225580215454102,
+      "learning_rate": 4.202202202202202e-06,
+      "loss": 0.3875,
+      "step": 57920
+    },
+    {
+      "epoch": 173.96,
+      "grad_norm": 12.730416297912598,
+      "learning_rate": 4.201201201201202e-06,
+      "loss": 0.3858,
+      "step": 57930
+    },
+    {
+      "epoch": 173.99,
+      "grad_norm": 14.95688247680664,
+      "learning_rate": 4.2002002002002e-06,
+      "loss": 0.4086,
+      "step": 57940
+    },
+    {
+      "epoch": 174.0,
+      "eval_accuracy": 0.8656,
+      "eval_loss": 0.5240481495857239,
+      "eval_runtime": 12.7304,
+      "eval_samples_per_second": 785.521,
+      "eval_steps_per_second": 3.142,
+      "step": 57942
+    },
+    {
+      "epoch": 174.02,
+      "grad_norm": 12.397719383239746,
+      "learning_rate": 4.199199199199199e-06,
+      "loss": 0.378,
+      "step": 57950
+    },
+    {
+      "epoch": 174.05,
+      "grad_norm": 13.01889705657959,
+      "learning_rate": 4.1981981981981984e-06,
+      "loss": 0.3818,
+      "step": 57960
+    },
+    {
+      "epoch": 174.08,
+      "grad_norm": 15.69823169708252,
+      "learning_rate": 4.197197197197198e-06,
+      "loss": 0.3725,
+      "step": 57970
+    },
+    {
+      "epoch": 174.11,
+      "grad_norm": 16.906587600708008,
+      "learning_rate": 4.1961961961961965e-06,
+      "loss": 0.4284,
+      "step": 57980
+    },
+    {
+      "epoch": 174.14,
+      "grad_norm": 24.786605834960938,
+      "learning_rate": 4.195195195195196e-06,
+      "loss": 0.3627,
+      "step": 57990
+    },
+    {
+      "epoch": 174.17,
+      "grad_norm": 11.78819751739502,
+      "learning_rate": 4.1941941941941945e-06,
+      "loss": 0.35,
+      "step": 58000
+    },
+    {
+      "epoch": 174.2,
+      "grad_norm": 17.188108444213867,
+      "learning_rate": 4.193193193193193e-06,
+      "loss": 0.4538,
+      "step": 58010
+    },
+    {
+      "epoch": 174.23,
+      "grad_norm": 12.299708366394043,
+      "learning_rate": 4.1921921921921925e-06,
+      "loss": 0.3624,
+      "step": 58020
+    },
+    {
+      "epoch": 174.26,
+      "grad_norm": 10.943092346191406,
+      "learning_rate": 4.191191191191192e-06,
+      "loss": 0.3548,
+      "step": 58030
+    },
+    {
+      "epoch": 174.29,
+      "grad_norm": 15.271341323852539,
+      "learning_rate": 4.1901901901901906e-06,
+      "loss": 0.4087,
+      "step": 58040
+    },
+    {
+      "epoch": 174.32,
+      "grad_norm": 11.296847343444824,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.3455,
+      "step": 58050
+    },
+    {
+      "epoch": 174.35,
+      "grad_norm": 14.725099563598633,
+      "learning_rate": 4.188188188188189e-06,
+      "loss": 0.3443,
+      "step": 58060
+    },
+    {
+      "epoch": 174.38,
+      "grad_norm": 14.821683883666992,
+      "learning_rate": 4.187187187187188e-06,
+      "loss": 0.3864,
+      "step": 58070
+    },
+    {
+      "epoch": 174.41,
+      "grad_norm": 17.810293197631836,
+      "learning_rate": 4.186186186186187e-06,
+      "loss": 0.3744,
+      "step": 58080
+    },
+    {
+      "epoch": 174.44,
+      "grad_norm": 16.28727912902832,
+      "learning_rate": 4.185185185185185e-06,
+      "loss": 0.404,
+      "step": 58090
+    },
+    {
+      "epoch": 174.47,
+      "grad_norm": 14.769054412841797,
+      "learning_rate": 4.184184184184185e-06,
+      "loss": 0.3282,
+      "step": 58100
+    },
+    {
+      "epoch": 174.5,
+      "grad_norm": 20.041501998901367,
+      "learning_rate": 4.183183183183184e-06,
+      "loss": 0.354,
+      "step": 58110
+    },
+    {
+      "epoch": 174.53,
+      "grad_norm": 15.903398513793945,
+      "learning_rate": 4.182182182182183e-06,
+      "loss": 0.3579,
+      "step": 58120
+    },
+    {
+      "epoch": 174.56,
+      "grad_norm": 14.09316635131836,
+      "learning_rate": 4.181181181181181e-06,
+      "loss": 0.3318,
+      "step": 58130
+    },
+    {
+      "epoch": 174.59,
+      "grad_norm": 16.802202224731445,
+      "learning_rate": 4.180180180180181e-06,
+      "loss": 0.3499,
+      "step": 58140
+    },
+    {
+      "epoch": 174.62,
+      "grad_norm": 20.073360443115234,
+      "learning_rate": 4.179179179179179e-06,
+      "loss": 0.3913,
+      "step": 58150
+    },
+    {
+      "epoch": 174.65,
+      "grad_norm": 27.02880096435547,
+      "learning_rate": 4.178178178178178e-06,
+      "loss": 0.3782,
+      "step": 58160
+    },
+    {
+      "epoch": 174.68,
+      "grad_norm": 14.8309907913208,
+      "learning_rate": 4.177177177177177e-06,
+      "loss": 0.3823,
+      "step": 58170
+    },
+    {
+      "epoch": 174.71,
+      "grad_norm": 23.147708892822266,
+      "learning_rate": 4.176176176176177e-06,
+      "loss": 0.3653,
+      "step": 58180
+    },
+    {
+      "epoch": 174.74,
+      "grad_norm": 18.570674896240234,
+      "learning_rate": 4.175175175175175e-06,
+      "loss": 0.3642,
+      "step": 58190
+    },
+    {
+      "epoch": 174.77,
+      "grad_norm": 18.13680076599121,
+      "learning_rate": 4.174174174174174e-06,
+      "loss": 0.3867,
+      "step": 58200
+    },
+    {
+      "epoch": 174.8,
+      "grad_norm": 13.197561264038086,
+      "learning_rate": 4.1731731731731735e-06,
+      "loss": 0.3517,
+      "step": 58210
+    },
+    {
+      "epoch": 174.83,
+      "grad_norm": 14.932595252990723,
+      "learning_rate": 4.172172172172173e-06,
+      "loss": 0.4325,
+      "step": 58220
+    },
+    {
+      "epoch": 174.86,
+      "grad_norm": 19.63214111328125,
+      "learning_rate": 4.1711711711711715e-06,
+      "loss": 0.4392,
+      "step": 58230
+    },
+    {
+      "epoch": 174.89,
+      "grad_norm": 21.568466186523438,
+      "learning_rate": 4.170170170170171e-06,
+      "loss": 0.3889,
+      "step": 58240
+    },
+    {
+      "epoch": 174.92,
+      "grad_norm": 11.348767280578613,
+      "learning_rate": 4.1691691691691695e-06,
+      "loss": 0.3436,
+      "step": 58250
+    },
+    {
+      "epoch": 174.95,
+      "grad_norm": 15.653636932373047,
+      "learning_rate": 4.168168168168168e-06,
+      "loss": 0.4208,
+      "step": 58260
+    },
+    {
+      "epoch": 174.98,
+      "grad_norm": 25.25259780883789,
+      "learning_rate": 4.1671671671671676e-06,
+      "loss": 0.3832,
+      "step": 58270
+    },
+    {
+      "epoch": 175.0,
+      "eval_accuracy": 0.8631,
+      "eval_loss": 0.5280221104621887,
+      "eval_runtime": 12.3906,
+      "eval_samples_per_second": 807.066,
+      "eval_steps_per_second": 3.228,
+      "step": 58275
+    },
+    {
+      "epoch": 175.02,
+      "grad_norm": 18.095916748046875,
+      "learning_rate": 4.166166166166167e-06,
+      "loss": 0.3941,
+      "step": 58280
+    },
+    {
+      "epoch": 175.05,
+      "grad_norm": 27.822181701660156,
+      "learning_rate": 4.165165165165166e-06,
+      "loss": 0.3983,
+      "step": 58290
+    },
+    {
+      "epoch": 175.08,
+      "grad_norm": 25.956199645996094,
+      "learning_rate": 4.164164164164164e-06,
+      "loss": 0.3608,
+      "step": 58300
+    },
+    {
+      "epoch": 175.11,
+      "grad_norm": 14.407609939575195,
+      "learning_rate": 4.163163163163164e-06,
+      "loss": 0.3485,
+      "step": 58310
+    },
+    {
+      "epoch": 175.14,
+      "grad_norm": 18.04946517944336,
+      "learning_rate": 4.162162162162163e-06,
+      "loss": 0.3273,
+      "step": 58320
+    },
+    {
+      "epoch": 175.17,
+      "grad_norm": 12.385822296142578,
+      "learning_rate": 4.161161161161162e-06,
+      "loss": 0.3519,
+      "step": 58330
+    },
+    {
+      "epoch": 175.2,
+      "grad_norm": 14.810163497924805,
+      "learning_rate": 4.16016016016016e-06,
+      "loss": 0.3934,
+      "step": 58340
+    },
+    {
+      "epoch": 175.23,
+      "grad_norm": 12.821239471435547,
+      "learning_rate": 4.15915915915916e-06,
+      "loss": 0.3425,
+      "step": 58350
+    },
+    {
+      "epoch": 175.26,
+      "grad_norm": 13.793792724609375,
+      "learning_rate": 4.158158158158158e-06,
+      "loss": 0.3765,
+      "step": 58360
+    },
+    {
+      "epoch": 175.29,
+      "grad_norm": 19.431360244750977,
+      "learning_rate": 4.157157157157157e-06,
+      "loss": 0.394,
+      "step": 58370
+    },
+    {
+      "epoch": 175.32,
+      "grad_norm": 18.780597686767578,
+      "learning_rate": 4.156156156156156e-06,
+      "loss": 0.3651,
+      "step": 58380
+    },
+    {
+      "epoch": 175.35,
+      "grad_norm": 16.881032943725586,
+      "learning_rate": 4.155155155155156e-06,
+      "loss": 0.3682,
+      "step": 58390
+    },
+    {
+      "epoch": 175.38,
+      "grad_norm": 32.65627670288086,
+      "learning_rate": 4.154154154154154e-06,
+      "loss": 0.3836,
+      "step": 58400
+    },
+    {
+      "epoch": 175.41,
+      "grad_norm": 18.420740127563477,
+      "learning_rate": 4.153153153153153e-06,
+      "loss": 0.3246,
+      "step": 58410
+    },
+    {
+      "epoch": 175.44,
+      "grad_norm": 16.31403923034668,
+      "learning_rate": 4.152152152152152e-06,
+      "loss": 0.382,
+      "step": 58420
+    },
+    {
+      "epoch": 175.47,
+      "grad_norm": 15.984549522399902,
+      "learning_rate": 4.151151151151152e-06,
+      "loss": 0.3814,
+      "step": 58430
+    },
+    {
+      "epoch": 175.5,
+      "grad_norm": 17.84511375427246,
+      "learning_rate": 4.1501501501501505e-06,
+      "loss": 0.3461,
+      "step": 58440
+    },
+    {
+      "epoch": 175.53,
+      "grad_norm": 26.348417282104492,
+      "learning_rate": 4.149149149149149e-06,
+      "loss": 0.364,
+      "step": 58450
+    },
+    {
+      "epoch": 175.56,
+      "grad_norm": 14.13006591796875,
+      "learning_rate": 4.1481481481481485e-06,
+      "loss": 0.3718,
+      "step": 58460
+    },
+    {
+      "epoch": 175.59,
+      "grad_norm": 15.95889663696289,
+      "learning_rate": 4.147147147147147e-06,
+      "loss": 0.3552,
+      "step": 58470
+    },
+    {
+      "epoch": 175.62,
+      "grad_norm": 22.39087677001953,
+      "learning_rate": 4.1461461461461465e-06,
+      "loss": 0.4174,
+      "step": 58480
+    },
+    {
+      "epoch": 175.65,
+      "grad_norm": 21.898983001708984,
+      "learning_rate": 4.145145145145145e-06,
+      "loss": 0.3967,
+      "step": 58490
+    },
+    {
+      "epoch": 175.68,
+      "grad_norm": 22.50005340576172,
+      "learning_rate": 4.1441441441441446e-06,
+      "loss": 0.412,
+      "step": 58500
+    },
+    {
+      "epoch": 175.71,
+      "grad_norm": 19.305355072021484,
+      "learning_rate": 4.143143143143143e-06,
+      "loss": 0.3702,
+      "step": 58510
+    },
+    {
+      "epoch": 175.74,
+      "grad_norm": 14.717147827148438,
+      "learning_rate": 4.142142142142143e-06,
+      "loss": 0.3508,
+      "step": 58520
+    },
+    {
+      "epoch": 175.77,
+      "grad_norm": 12.45810317993164,
+      "learning_rate": 4.141141141141142e-06,
+      "loss": 0.3253,
+      "step": 58530
+    },
+    {
+      "epoch": 175.8,
+      "grad_norm": 13.548640251159668,
+      "learning_rate": 4.140140140140141e-06,
+      "loss": 0.4433,
+      "step": 58540
+    },
+    {
+      "epoch": 175.83,
+      "grad_norm": 15.92135238647461,
+      "learning_rate": 4.139139139139139e-06,
+      "loss": 0.3569,
+      "step": 58550
+    },
+    {
+      "epoch": 175.86,
+      "grad_norm": 15.379179954528809,
+      "learning_rate": 4.138138138138139e-06,
+      "loss": 0.3609,
+      "step": 58560
+    },
+    {
+      "epoch": 175.89,
+      "grad_norm": 14.361334800720215,
+      "learning_rate": 4.137137137137138e-06,
+      "loss": 0.4164,
+      "step": 58570
+    },
+    {
+      "epoch": 175.92,
+      "grad_norm": 15.73035717010498,
+      "learning_rate": 4.136136136136137e-06,
+      "loss": 0.4167,
+      "step": 58580
+    },
+    {
+      "epoch": 175.95,
+      "grad_norm": 12.671379089355469,
+      "learning_rate": 4.135135135135135e-06,
+      "loss": 0.3984,
+      "step": 58590
+    },
+    {
+      "epoch": 175.98,
+      "grad_norm": 14.838611602783203,
+      "learning_rate": 4.134134134134135e-06,
+      "loss": 0.3515,
+      "step": 58600
+    },
+    {
+      "epoch": 176.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5301606059074402,
+      "eval_runtime": 12.6649,
+      "eval_samples_per_second": 789.581,
+      "eval_steps_per_second": 3.158,
+      "step": 58608
+    },
+    {
+      "epoch": 176.01,
+      "grad_norm": 36.09259796142578,
+      "learning_rate": 4.133133133133133e-06,
+      "loss": 0.362,
+      "step": 58610
+    },
+    {
+      "epoch": 176.04,
+      "grad_norm": 14.761162757873535,
+      "learning_rate": 4.132132132132132e-06,
+      "loss": 0.3916,
+      "step": 58620
+    },
+    {
+      "epoch": 176.07,
+      "grad_norm": 15.46599292755127,
+      "learning_rate": 4.131131131131131e-06,
+      "loss": 0.382,
+      "step": 58630
+    },
+    {
+      "epoch": 176.1,
+      "grad_norm": 23.60271453857422,
+      "learning_rate": 4.130130130130131e-06,
+      "loss": 0.3477,
+      "step": 58640
+    },
+    {
+      "epoch": 176.13,
+      "grad_norm": 14.518853187561035,
+      "learning_rate": 4.129129129129129e-06,
+      "loss": 0.3765,
+      "step": 58650
+    },
+    {
+      "epoch": 176.16,
+      "grad_norm": 26.77389907836914,
+      "learning_rate": 4.128128128128128e-06,
+      "loss": 0.3436,
+      "step": 58660
+    },
+    {
+      "epoch": 176.19,
+      "grad_norm": 13.1908597946167,
+      "learning_rate": 4.1271271271271275e-06,
+      "loss": 0.4021,
+      "step": 58670
+    },
+    {
+      "epoch": 176.22,
+      "grad_norm": 14.42405891418457,
+      "learning_rate": 4.126126126126127e-06,
+      "loss": 0.3916,
+      "step": 58680
+    },
+    {
+      "epoch": 176.25,
+      "grad_norm": 15.440362930297852,
+      "learning_rate": 4.1251251251251255e-06,
+      "loss": 0.3776,
+      "step": 58690
+    },
+    {
+      "epoch": 176.28,
+      "grad_norm": 22.7329044342041,
+      "learning_rate": 4.124124124124124e-06,
+      "loss": 0.3614,
+      "step": 58700
+    },
+    {
+      "epoch": 176.31,
+      "grad_norm": 18.822298049926758,
+      "learning_rate": 4.1231231231231235e-06,
+      "loss": 0.3607,
+      "step": 58710
+    },
+    {
+      "epoch": 176.34,
+      "grad_norm": 17.130443572998047,
+      "learning_rate": 4.122122122122122e-06,
+      "loss": 0.4009,
+      "step": 58720
+    },
+    {
+      "epoch": 176.37,
+      "grad_norm": 35.27191925048828,
+      "learning_rate": 4.1211211211211216e-06,
+      "loss": 0.3663,
+      "step": 58730
+    },
+    {
+      "epoch": 176.4,
+      "grad_norm": 15.46333122253418,
+      "learning_rate": 4.12012012012012e-06,
+      "loss": 0.3335,
+      "step": 58740
+    },
+    {
+      "epoch": 176.43,
+      "grad_norm": 25.11519432067871,
+      "learning_rate": 4.11911911911912e-06,
+      "loss": 0.3463,
+      "step": 58750
+    },
+    {
+      "epoch": 176.46,
+      "grad_norm": 21.27839469909668,
+      "learning_rate": 4.118118118118118e-06,
+      "loss": 0.3256,
+      "step": 58760
+    },
+    {
+      "epoch": 176.49,
+      "grad_norm": 18.217618942260742,
+      "learning_rate": 4.117117117117118e-06,
+      "loss": 0.386,
+      "step": 58770
+    },
+    {
+      "epoch": 176.52,
+      "grad_norm": 16.21112632751465,
+      "learning_rate": 4.116116116116117e-06,
+      "loss": 0.3778,
+      "step": 58780
+    },
+    {
+      "epoch": 176.55,
+      "grad_norm": 17.2326602935791,
+      "learning_rate": 4.115115115115116e-06,
+      "loss": 0.3981,
+      "step": 58790
+    },
+    {
+      "epoch": 176.58,
+      "grad_norm": 12.765154838562012,
+      "learning_rate": 4.114114114114114e-06,
+      "loss": 0.3271,
+      "step": 58800
+    },
+    {
+      "epoch": 176.61,
+      "grad_norm": 25.723731994628906,
+      "learning_rate": 4.113113113113114e-06,
+      "loss": 0.377,
+      "step": 58810
+    },
+    {
+      "epoch": 176.64,
+      "grad_norm": 13.36816692352295,
+      "learning_rate": 4.112112112112112e-06,
+      "loss": 0.312,
+      "step": 58820
+    },
+    {
+      "epoch": 176.67,
+      "grad_norm": 27.906890869140625,
+      "learning_rate": 4.111111111111111e-06,
+      "loss": 0.3799,
+      "step": 58830
+    },
+    {
+      "epoch": 176.7,
+      "grad_norm": 10.003141403198242,
+      "learning_rate": 4.11011011011011e-06,
+      "loss": 0.3682,
+      "step": 58840
+    },
+    {
+      "epoch": 176.73,
+      "grad_norm": 21.49915885925293,
+      "learning_rate": 4.10910910910911e-06,
+      "loss": 0.354,
+      "step": 58850
+    },
+    {
+      "epoch": 176.76,
+      "grad_norm": 18.625463485717773,
+      "learning_rate": 4.108108108108108e-06,
+      "loss": 0.3612,
+      "step": 58860
+    },
+    {
+      "epoch": 176.79,
+      "grad_norm": 13.837291717529297,
+      "learning_rate": 4.107107107107107e-06,
+      "loss": 0.3917,
+      "step": 58870
+    },
+    {
+      "epoch": 176.82,
+      "grad_norm": 15.597818374633789,
+      "learning_rate": 4.106106106106106e-06,
+      "loss": 0.3669,
+      "step": 58880
+    },
+    {
+      "epoch": 176.85,
+      "grad_norm": 17.365896224975586,
+      "learning_rate": 4.105105105105106e-06,
+      "loss": 0.4072,
+      "step": 58890
+    },
+    {
+      "epoch": 176.88,
+      "grad_norm": 14.935830116271973,
+      "learning_rate": 4.1041041041041045e-06,
+      "loss": 0.3604,
+      "step": 58900
+    },
+    {
+      "epoch": 176.91,
+      "grad_norm": 17.3485107421875,
+      "learning_rate": 4.103103103103103e-06,
+      "loss": 0.3934,
+      "step": 58910
+    },
+    {
+      "epoch": 176.94,
+      "grad_norm": 17.555707931518555,
+      "learning_rate": 4.1021021021021025e-06,
+      "loss": 0.3665,
+      "step": 58920
+    },
+    {
+      "epoch": 176.97,
+      "grad_norm": 12.081335067749023,
+      "learning_rate": 4.101101101101101e-06,
+      "loss": 0.3355,
+      "step": 58930
+    },
+    {
+      "epoch": 177.0,
+      "grad_norm": 12.608793258666992,
+      "learning_rate": 4.1001001001001005e-06,
+      "loss": 0.3749,
+      "step": 58940
+    },
+    {
+      "epoch": 177.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.531607985496521,
+      "eval_runtime": 12.8584,
+      "eval_samples_per_second": 777.7,
+      "eval_steps_per_second": 3.111,
+      "step": 58941
+    },
+    {
+      "epoch": 177.03,
+      "grad_norm": 24.641517639160156,
+      "learning_rate": 4.099099099099099e-06,
+      "loss": 0.3276,
+      "step": 58950
+    },
+    {
+      "epoch": 177.06,
+      "grad_norm": 14.71824836730957,
+      "learning_rate": 4.0980980980980986e-06,
+      "loss": 0.3363,
+      "step": 58960
+    },
+    {
+      "epoch": 177.09,
+      "grad_norm": 18.64984130859375,
+      "learning_rate": 4.097097097097097e-06,
+      "loss": 0.4086,
+      "step": 58970
+    },
+    {
+      "epoch": 177.12,
+      "grad_norm": 14.29690933227539,
+      "learning_rate": 4.096096096096097e-06,
+      "loss": 0.3641,
+      "step": 58980
+    },
+    {
+      "epoch": 177.15,
+      "grad_norm": 11.251493453979492,
+      "learning_rate": 4.095095095095095e-06,
+      "loss": 0.3207,
+      "step": 58990
+    },
+    {
+      "epoch": 177.18,
+      "grad_norm": 16.051326751708984,
+      "learning_rate": 4.094094094094095e-06,
+      "loss": 0.3181,
+      "step": 59000
+    },
+    {
+      "epoch": 177.21,
+      "grad_norm": 17.022438049316406,
+      "learning_rate": 4.093093093093093e-06,
+      "loss": 0.3392,
+      "step": 59010
+    },
+    {
+      "epoch": 177.24,
+      "grad_norm": 13.676738739013672,
+      "learning_rate": 4.092092092092093e-06,
+      "loss": 0.3652,
+      "step": 59020
+    },
+    {
+      "epoch": 177.27,
+      "grad_norm": 14.524504661560059,
+      "learning_rate": 4.091091091091091e-06,
+      "loss": 0.3849,
+      "step": 59030
+    },
+    {
+      "epoch": 177.3,
+      "grad_norm": 20.60028076171875,
+      "learning_rate": 4.09009009009009e-06,
+      "loss": 0.3474,
+      "step": 59040
+    },
+    {
+      "epoch": 177.33,
+      "grad_norm": 15.16180419921875,
+      "learning_rate": 4.089089089089089e-06,
+      "loss": 0.318,
+      "step": 59050
+    },
+    {
+      "epoch": 177.36,
+      "grad_norm": 14.023417472839355,
+      "learning_rate": 4.088088088088089e-06,
+      "loss": 0.3382,
+      "step": 59060
+    },
+    {
+      "epoch": 177.39,
+      "grad_norm": 10.815349578857422,
+      "learning_rate": 4.087087087087087e-06,
+      "loss": 0.3633,
+      "step": 59070
+    },
+    {
+      "epoch": 177.42,
+      "grad_norm": 14.608118057250977,
+      "learning_rate": 4.086086086086086e-06,
+      "loss": 0.4097,
+      "step": 59080
+    },
+    {
+      "epoch": 177.45,
+      "grad_norm": 15.711928367614746,
+      "learning_rate": 4.085085085085085e-06,
+      "loss": 0.3754,
+      "step": 59090
+    },
+    {
+      "epoch": 177.48,
+      "grad_norm": 13.490781784057617,
+      "learning_rate": 4.084084084084085e-06,
+      "loss": 0.3659,
+      "step": 59100
+    },
+    {
+      "epoch": 177.51,
+      "grad_norm": 11.90878963470459,
+      "learning_rate": 4.083083083083083e-06,
+      "loss": 0.3281,
+      "step": 59110
+    },
+    {
+      "epoch": 177.54,
+      "grad_norm": 21.53923988342285,
+      "learning_rate": 4.082082082082082e-06,
+      "loss": 0.3921,
+      "step": 59120
+    },
+    {
+      "epoch": 177.57,
+      "grad_norm": 15.465372085571289,
+      "learning_rate": 4.0810810810810815e-06,
+      "loss": 0.3528,
+      "step": 59130
+    },
+    {
+      "epoch": 177.6,
+      "grad_norm": 15.061660766601562,
+      "learning_rate": 4.080080080080081e-06,
+      "loss": 0.3581,
+      "step": 59140
+    },
+    {
+      "epoch": 177.63,
+      "grad_norm": 7.823965072631836,
+      "learning_rate": 4.0790790790790795e-06,
+      "loss": 0.3383,
+      "step": 59150
+    },
+    {
+      "epoch": 177.66,
+      "grad_norm": 13.462827682495117,
+      "learning_rate": 4.078078078078078e-06,
+      "loss": 0.327,
+      "step": 59160
+    },
+    {
+      "epoch": 177.69,
+      "grad_norm": 25.48688316345215,
+      "learning_rate": 4.0770770770770775e-06,
+      "loss": 0.403,
+      "step": 59170
+    },
+    {
+      "epoch": 177.72,
+      "grad_norm": 17.938016891479492,
+      "learning_rate": 4.076076076076076e-06,
+      "loss": 0.3824,
+      "step": 59180
+    },
+    {
+      "epoch": 177.75,
+      "grad_norm": 21.605228424072266,
+      "learning_rate": 4.0750750750750756e-06,
+      "loss": 0.4189,
+      "step": 59190
+    },
+    {
+      "epoch": 177.78,
+      "grad_norm": 16.73281478881836,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.3402,
+      "step": 59200
+    },
+    {
+      "epoch": 177.81,
+      "grad_norm": 17.04777717590332,
+      "learning_rate": 4.073073073073074e-06,
+      "loss": 0.355,
+      "step": 59210
+    },
+    {
+      "epoch": 177.84,
+      "grad_norm": 19.468130111694336,
+      "learning_rate": 4.072072072072072e-06,
+      "loss": 0.3869,
+      "step": 59220
+    },
+    {
+      "epoch": 177.87,
+      "grad_norm": 14.405555725097656,
+      "learning_rate": 4.071071071071072e-06,
+      "loss": 0.3443,
+      "step": 59230
+    },
+    {
+      "epoch": 177.9,
+      "grad_norm": 19.33612632751465,
+      "learning_rate": 4.07007007007007e-06,
+      "loss": 0.3906,
+      "step": 59240
+    },
+    {
+      "epoch": 177.93,
+      "grad_norm": 13.25005054473877,
+      "learning_rate": 4.06906906906907e-06,
+      "loss": 0.326,
+      "step": 59250
+    },
+    {
+      "epoch": 177.96,
+      "grad_norm": 15.196127891540527,
+      "learning_rate": 4.068068068068068e-06,
+      "loss": 0.4157,
+      "step": 59260
+    },
+    {
+      "epoch": 177.99,
+      "grad_norm": 17.374542236328125,
+      "learning_rate": 4.067067067067068e-06,
+      "loss": 0.3298,
+      "step": 59270
+    },
+    {
+      "epoch": 178.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.5290198922157288,
+      "eval_runtime": 13.2587,
+      "eval_samples_per_second": 754.22,
+      "eval_steps_per_second": 3.017,
+      "step": 59274
+    },
+    {
+      "epoch": 178.02,
+      "grad_norm": 10.847558975219727,
+      "learning_rate": 4.066066066066066e-06,
+      "loss": 0.3826,
+      "step": 59280
+    },
+    {
+      "epoch": 178.05,
+      "grad_norm": 19.37006378173828,
+      "learning_rate": 4.065065065065065e-06,
+      "loss": 0.3704,
+      "step": 59290
+    },
+    {
+      "epoch": 178.08,
+      "grad_norm": 17.6607608795166,
+      "learning_rate": 4.064064064064064e-06,
+      "loss": 0.345,
+      "step": 59300
+    },
+    {
+      "epoch": 178.11,
+      "grad_norm": 21.62128257751465,
+      "learning_rate": 4.063063063063064e-06,
+      "loss": 0.3725,
+      "step": 59310
+    },
+    {
+      "epoch": 178.14,
+      "grad_norm": 15.364421844482422,
+      "learning_rate": 4.062062062062062e-06,
+      "loss": 0.3221,
+      "step": 59320
+    },
+    {
+      "epoch": 178.17,
+      "grad_norm": 17.84992027282715,
+      "learning_rate": 4.061061061061061e-06,
+      "loss": 0.3152,
+      "step": 59330
+    },
+    {
+      "epoch": 178.2,
+      "grad_norm": 15.065093040466309,
+      "learning_rate": 4.06006006006006e-06,
+      "loss": 0.3637,
+      "step": 59340
+    },
+    {
+      "epoch": 178.23,
+      "grad_norm": 9.66307258605957,
+      "learning_rate": 4.05905905905906e-06,
+      "loss": 0.4044,
+      "step": 59350
+    },
+    {
+      "epoch": 178.26,
+      "grad_norm": 20.396486282348633,
+      "learning_rate": 4.0580580580580584e-06,
+      "loss": 0.3907,
+      "step": 59360
+    },
+    {
+      "epoch": 178.29,
+      "grad_norm": 28.590564727783203,
+      "learning_rate": 4.057057057057057e-06,
+      "loss": 0.3799,
+      "step": 59370
+    },
+    {
+      "epoch": 178.32,
+      "grad_norm": 19.778085708618164,
+      "learning_rate": 4.0560560560560565e-06,
+      "loss": 0.4137,
+      "step": 59380
+    },
+    {
+      "epoch": 178.35,
+      "grad_norm": 14.239270210266113,
+      "learning_rate": 4.055055055055055e-06,
+      "loss": 0.3314,
+      "step": 59390
+    },
+    {
+      "epoch": 178.38,
+      "grad_norm": 12.554594039916992,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.3686,
+      "step": 59400
+    },
+    {
+      "epoch": 178.41,
+      "grad_norm": 17.125394821166992,
+      "learning_rate": 4.053053053053053e-06,
+      "loss": 0.3657,
+      "step": 59410
+    },
+    {
+      "epoch": 178.44,
+      "grad_norm": 15.225324630737305,
+      "learning_rate": 4.0520520520520526e-06,
+      "loss": 0.3963,
+      "step": 59420
+    },
+    {
+      "epoch": 178.47,
+      "grad_norm": 10.903555870056152,
+      "learning_rate": 4.051051051051051e-06,
+      "loss": 0.3836,
+      "step": 59430
+    },
+    {
+      "epoch": 178.5,
+      "grad_norm": 11.631128311157227,
+      "learning_rate": 4.050050050050051e-06,
+      "loss": 0.3701,
+      "step": 59440
+    },
+    {
+      "epoch": 178.53,
+      "grad_norm": 15.767228126525879,
+      "learning_rate": 4.049049049049049e-06,
+      "loss": 0.4017,
+      "step": 59450
+    },
+    {
+      "epoch": 178.56,
+      "grad_norm": 18.55630874633789,
+      "learning_rate": 4.048048048048049e-06,
+      "loss": 0.4052,
+      "step": 59460
+    },
+    {
+      "epoch": 178.59,
+      "grad_norm": 17.0009765625,
+      "learning_rate": 4.047047047047047e-06,
+      "loss": 0.3619,
+      "step": 59470
+    },
+    {
+      "epoch": 178.62,
+      "grad_norm": 14.285255432128906,
+      "learning_rate": 4.046046046046047e-06,
+      "loss": 0.3595,
+      "step": 59480
+    },
+    {
+      "epoch": 178.65,
+      "grad_norm": 14.716629981994629,
+      "learning_rate": 4.045045045045045e-06,
+      "loss": 0.3532,
+      "step": 59490
+    },
+    {
+      "epoch": 178.68,
+      "grad_norm": 18.379032135009766,
+      "learning_rate": 4.044044044044044e-06,
+      "loss": 0.3308,
+      "step": 59500
+    },
+    {
+      "epoch": 178.71,
+      "grad_norm": 15.98036003112793,
+      "learning_rate": 4.043043043043043e-06,
+      "loss": 0.3812,
+      "step": 59510
+    },
+    {
+      "epoch": 178.74,
+      "grad_norm": 13.0855073928833,
+      "learning_rate": 4.042042042042043e-06,
+      "loss": 0.3616,
+      "step": 59520
+    },
+    {
+      "epoch": 178.77,
+      "grad_norm": 14.513360977172852,
+      "learning_rate": 4.041041041041041e-06,
+      "loss": 0.3586,
+      "step": 59530
+    },
+    {
+      "epoch": 178.8,
+      "grad_norm": 15.116728782653809,
+      "learning_rate": 4.04004004004004e-06,
+      "loss": 0.3708,
+      "step": 59540
+    },
+    {
+      "epoch": 178.83,
+      "grad_norm": 20.748456954956055,
+      "learning_rate": 4.039039039039039e-06,
+      "loss": 0.3839,
+      "step": 59550
+    },
+    {
+      "epoch": 178.86,
+      "grad_norm": 28.844690322875977,
+      "learning_rate": 4.038038038038039e-06,
+      "loss": 0.4153,
+      "step": 59560
+    },
+    {
+      "epoch": 178.89,
+      "grad_norm": 23.72994041442871,
+      "learning_rate": 4.037037037037037e-06,
+      "loss": 0.4186,
+      "step": 59570
+    },
+    {
+      "epoch": 178.92,
+      "grad_norm": 19.090591430664062,
+      "learning_rate": 4.036036036036036e-06,
+      "loss": 0.3662,
+      "step": 59580
+    },
+    {
+      "epoch": 178.95,
+      "grad_norm": 14.53249740600586,
+      "learning_rate": 4.0350350350350354e-06,
+      "loss": 0.3569,
+      "step": 59590
+    },
+    {
+      "epoch": 178.98,
+      "grad_norm": 14.516371726989746,
+      "learning_rate": 4.034034034034035e-06,
+      "loss": 0.3758,
+      "step": 59600
+    },
+    {
+      "epoch": 179.0,
+      "eval_accuracy": 0.8668,
+      "eval_loss": 0.5272110104560852,
+      "eval_runtime": 12.907,
+      "eval_samples_per_second": 774.775,
+      "eval_steps_per_second": 3.099,
+      "step": 59607
+    },
+    {
+      "epoch": 179.01,
+      "grad_norm": 14.402239799499512,
+      "learning_rate": 4.0330330330330335e-06,
+      "loss": 0.3759,
+      "step": 59610
+    },
+    {
+      "epoch": 179.04,
+      "grad_norm": 16.124462127685547,
+      "learning_rate": 4.032032032032032e-06,
+      "loss": 0.4043,
+      "step": 59620
+    },
+    {
+      "epoch": 179.07,
+      "grad_norm": 13.552022933959961,
+      "learning_rate": 4.0310310310310315e-06,
+      "loss": 0.3843,
+      "step": 59630
+    },
+    {
+      "epoch": 179.1,
+      "grad_norm": 11.932546615600586,
+      "learning_rate": 4.03003003003003e-06,
+      "loss": 0.3324,
+      "step": 59640
+    },
+    {
+      "epoch": 179.13,
+      "grad_norm": 15.692938804626465,
+      "learning_rate": 4.0290290290290296e-06,
+      "loss": 0.3857,
+      "step": 59650
+    },
+    {
+      "epoch": 179.16,
+      "grad_norm": 18.935623168945312,
+      "learning_rate": 4.028028028028028e-06,
+      "loss": 0.3869,
+      "step": 59660
+    },
+    {
+      "epoch": 179.19,
+      "grad_norm": 16.247264862060547,
+      "learning_rate": 4.027027027027028e-06,
+      "loss": 0.3686,
+      "step": 59670
+    },
+    {
+      "epoch": 179.22,
+      "grad_norm": 16.911930084228516,
+      "learning_rate": 4.026026026026026e-06,
+      "loss": 0.4199,
+      "step": 59680
+    },
+    {
+      "epoch": 179.25,
+      "grad_norm": 11.424375534057617,
+      "learning_rate": 4.025025025025026e-06,
+      "loss": 0.3819,
+      "step": 59690
+    },
+    {
+      "epoch": 179.28,
+      "grad_norm": 17.458454132080078,
+      "learning_rate": 4.024024024024024e-06,
+      "loss": 0.3882,
+      "step": 59700
+    },
+    {
+      "epoch": 179.31,
+      "grad_norm": 16.669322967529297,
+      "learning_rate": 4.023023023023024e-06,
+      "loss": 0.3877,
+      "step": 59710
+    },
+    {
+      "epoch": 179.34,
+      "grad_norm": 20.44898223876953,
+      "learning_rate": 4.022022022022022e-06,
+      "loss": 0.4004,
+      "step": 59720
+    },
+    {
+      "epoch": 179.37,
+      "grad_norm": 19.079898834228516,
+      "learning_rate": 4.021021021021022e-06,
+      "loss": 0.3623,
+      "step": 59730
+    },
+    {
+      "epoch": 179.4,
+      "grad_norm": 13.059213638305664,
+      "learning_rate": 4.02002002002002e-06,
+      "loss": 0.4243,
+      "step": 59740
+    },
+    {
+      "epoch": 179.43,
+      "grad_norm": 21.410655975341797,
+      "learning_rate": 4.019019019019019e-06,
+      "loss": 0.3836,
+      "step": 59750
+    },
+    {
+      "epoch": 179.46,
+      "grad_norm": 14.035740852355957,
+      "learning_rate": 4.018018018018018e-06,
+      "loss": 0.4113,
+      "step": 59760
+    },
+    {
+      "epoch": 179.49,
+      "grad_norm": 17.12101936340332,
+      "learning_rate": 4.017017017017018e-06,
+      "loss": 0.3387,
+      "step": 59770
+    },
+    {
+      "epoch": 179.52,
+      "grad_norm": 15.087050437927246,
+      "learning_rate": 4.016016016016016e-06,
+      "loss": 0.3685,
+      "step": 59780
+    },
+    {
+      "epoch": 179.55,
+      "grad_norm": 16.536413192749023,
+      "learning_rate": 4.015015015015015e-06,
+      "loss": 0.3541,
+      "step": 59790
+    },
+    {
+      "epoch": 179.58,
+      "grad_norm": 27.538612365722656,
+      "learning_rate": 4.014014014014014e-06,
+      "loss": 0.4022,
+      "step": 59800
+    },
+    {
+      "epoch": 179.61,
+      "grad_norm": 11.606536865234375,
+      "learning_rate": 4.013013013013014e-06,
+      "loss": 0.3444,
+      "step": 59810
+    },
+    {
+      "epoch": 179.64,
+      "grad_norm": 12.86123275756836,
+      "learning_rate": 4.0120120120120124e-06,
+      "loss": 0.2946,
+      "step": 59820
+    },
+    {
+      "epoch": 179.67,
+      "grad_norm": 15.77061653137207,
+      "learning_rate": 4.011011011011011e-06,
+      "loss": 0.3734,
+      "step": 59830
+    },
+    {
+      "epoch": 179.7,
+      "grad_norm": 17.93335723876953,
+      "learning_rate": 4.0100100100100105e-06,
+      "loss": 0.3686,
+      "step": 59840
+    },
+    {
+      "epoch": 179.73,
+      "grad_norm": 16.249404907226562,
+      "learning_rate": 4.009009009009009e-06,
+      "loss": 0.3863,
+      "step": 59850
+    },
+    {
+      "epoch": 179.76,
+      "grad_norm": 11.17872142791748,
+      "learning_rate": 4.0080080080080085e-06,
+      "loss": 0.3871,
+      "step": 59860
+    },
+    {
+      "epoch": 179.79,
+      "grad_norm": 20.134281158447266,
+      "learning_rate": 4.007007007007007e-06,
+      "loss": 0.382,
+      "step": 59870
+    },
+    {
+      "epoch": 179.82,
+      "grad_norm": 14.958374977111816,
+      "learning_rate": 4.0060060060060066e-06,
+      "loss": 0.3737,
+      "step": 59880
+    },
+    {
+      "epoch": 179.85,
+      "grad_norm": 17.80064582824707,
+      "learning_rate": 4.005005005005005e-06,
+      "loss": 0.4,
+      "step": 59890
+    },
+    {
+      "epoch": 179.88,
+      "grad_norm": 19.82135581970215,
+      "learning_rate": 4.004004004004005e-06,
+      "loss": 0.3633,
+      "step": 59900
+    },
+    {
+      "epoch": 179.91,
+      "grad_norm": 13.035211563110352,
+      "learning_rate": 4.003003003003003e-06,
+      "loss": 0.3619,
+      "step": 59910
+    },
+    {
+      "epoch": 179.94,
+      "grad_norm": 27.069429397583008,
+      "learning_rate": 4.002002002002003e-06,
+      "loss": 0.3789,
+      "step": 59920
+    },
+    {
+      "epoch": 179.97,
+      "grad_norm": 17.16834259033203,
+      "learning_rate": 4.001001001001001e-06,
+      "loss": 0.4071,
+      "step": 59930
+    },
+    {
+      "epoch": 180.0,
+      "grad_norm": 140.3048553466797,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.31,
+      "step": 59940
+    },
+    {
+      "epoch": 180.0,
+      "eval_accuracy": 0.864,
+      "eval_loss": 0.5314058661460876,
+      "eval_runtime": 12.6247,
+      "eval_samples_per_second": 792.1,
+      "eval_steps_per_second": 3.168,
+      "step": 59940
+    },
+    {
+      "epoch": 180.03,
+      "grad_norm": 17.123760223388672,
+      "learning_rate": 3.998998998998999e-06,
+      "loss": 0.3707,
+      "step": 59950
+    },
+    {
+      "epoch": 180.06,
+      "grad_norm": 16.074031829833984,
+      "learning_rate": 3.997997997997998e-06,
+      "loss": 0.3951,
+      "step": 59960
+    },
+    {
+      "epoch": 180.09,
+      "grad_norm": 16.777767181396484,
+      "learning_rate": 3.996996996996997e-06,
+      "loss": 0.3783,
+      "step": 59970
+    },
+    {
+      "epoch": 180.12,
+      "grad_norm": 13.733888626098633,
+      "learning_rate": 3.995995995995997e-06,
+      "loss": 0.3834,
+      "step": 59980
+    },
+    {
+      "epoch": 180.15,
+      "grad_norm": 16.989789962768555,
+      "learning_rate": 3.994994994994995e-06,
+      "loss": 0.3635,
+      "step": 59990
+    },
+    {
+      "epoch": 180.18,
+      "grad_norm": 34.987281799316406,
+      "learning_rate": 3.993993993993994e-06,
+      "loss": 0.3833,
+      "step": 60000
+    },
+    {
+      "epoch": 180.21,
+      "grad_norm": 20.086925506591797,
+      "learning_rate": 3.992992992992993e-06,
+      "loss": 0.3716,
+      "step": 60010
+    },
+    {
+      "epoch": 180.24,
+      "grad_norm": 11.589064598083496,
+      "learning_rate": 3.991991991991993e-06,
+      "loss": 0.3632,
+      "step": 60020
+    },
+    {
+      "epoch": 180.27,
+      "grad_norm": 19.051618576049805,
+      "learning_rate": 3.990990990990991e-06,
+      "loss": 0.3622,
+      "step": 60030
+    },
+    {
+      "epoch": 180.3,
+      "grad_norm": 12.951354026794434,
+      "learning_rate": 3.98998998998999e-06,
+      "loss": 0.4244,
+      "step": 60040
+    },
+    {
+      "epoch": 180.33,
+      "grad_norm": 18.54038429260254,
+      "learning_rate": 3.9889889889889894e-06,
+      "loss": 0.3464,
+      "step": 60050
+    },
+    {
+      "epoch": 180.36,
+      "grad_norm": 19.544721603393555,
+      "learning_rate": 3.987987987987989e-06,
+      "loss": 0.3893,
+      "step": 60060
+    },
+    {
+      "epoch": 180.39,
+      "grad_norm": 15.213372230529785,
+      "learning_rate": 3.986986986986987e-06,
+      "loss": 0.3698,
+      "step": 60070
+    },
+    {
+      "epoch": 180.42,
+      "grad_norm": 12.69859790802002,
+      "learning_rate": 3.985985985985986e-06,
+      "loss": 0.3643,
+      "step": 60080
+    },
+    {
+      "epoch": 180.45,
+      "grad_norm": 13.0181884765625,
+      "learning_rate": 3.9849849849849855e-06,
+      "loss": 0.3847,
+      "step": 60090
+    },
+    {
+      "epoch": 180.48,
+      "grad_norm": 20.600780487060547,
+      "learning_rate": 3.983983983983984e-06,
+      "loss": 0.3509,
+      "step": 60100
+    },
+    {
+      "epoch": 180.51,
+      "grad_norm": 15.574991226196289,
+      "learning_rate": 3.9829829829829836e-06,
+      "loss": 0.4063,
+      "step": 60110
+    },
+    {
+      "epoch": 180.54,
+      "grad_norm": 13.062969207763672,
+      "learning_rate": 3.981981981981982e-06,
+      "loss": 0.3951,
+      "step": 60120
+    },
+    {
+      "epoch": 180.57,
+      "grad_norm": 14.166492462158203,
+      "learning_rate": 3.980980980980982e-06,
+      "loss": 0.3289,
+      "step": 60130
+    },
+    {
+      "epoch": 180.6,
+      "grad_norm": 13.742244720458984,
+      "learning_rate": 3.97997997997998e-06,
+      "loss": 0.4083,
+      "step": 60140
+    },
+    {
+      "epoch": 180.63,
+      "grad_norm": 17.606794357299805,
+      "learning_rate": 3.97897897897898e-06,
+      "loss": 0.3783,
+      "step": 60150
+    },
+    {
+      "epoch": 180.66,
+      "grad_norm": 13.325634002685547,
+      "learning_rate": 3.977977977977978e-06,
+      "loss": 0.3749,
+      "step": 60160
+    },
+    {
+      "epoch": 180.69,
+      "grad_norm": 25.762454986572266,
+      "learning_rate": 3.976976976976978e-06,
+      "loss": 0.4004,
+      "step": 60170
+    },
+    {
+      "epoch": 180.72,
+      "grad_norm": 11.901933670043945,
+      "learning_rate": 3.975975975975976e-06,
+      "loss": 0.3423,
+      "step": 60180
+    },
+    {
+      "epoch": 180.75,
+      "grad_norm": 11.02151870727539,
+      "learning_rate": 3.974974974974976e-06,
+      "loss": 0.4062,
+      "step": 60190
+    },
+    {
+      "epoch": 180.78,
+      "grad_norm": 17.37810707092285,
+      "learning_rate": 3.973973973973974e-06,
+      "loss": 0.4239,
+      "step": 60200
+    },
+    {
+      "epoch": 180.81,
+      "grad_norm": 14.58978271484375,
+      "learning_rate": 3.972972972972973e-06,
+      "loss": 0.321,
+      "step": 60210
+    },
+    {
+      "epoch": 180.84,
+      "grad_norm": 7.685393810272217,
+      "learning_rate": 3.971971971971972e-06,
+      "loss": 0.3581,
+      "step": 60220
+    },
+    {
+      "epoch": 180.87,
+      "grad_norm": 19.577468872070312,
+      "learning_rate": 3.970970970970972e-06,
+      "loss": 0.3817,
+      "step": 60230
+    },
+    {
+      "epoch": 180.9,
+      "grad_norm": 16.966415405273438,
+      "learning_rate": 3.96996996996997e-06,
+      "loss": 0.3579,
+      "step": 60240
+    },
+    {
+      "epoch": 180.93,
+      "grad_norm": 25.40387535095215,
+      "learning_rate": 3.968968968968969e-06,
+      "loss": 0.3544,
+      "step": 60250
+    },
+    {
+      "epoch": 180.96,
+      "grad_norm": 14.105462074279785,
+      "learning_rate": 3.967967967967968e-06,
+      "loss": 0.377,
+      "step": 60260
+    },
+    {
+      "epoch": 180.99,
+      "grad_norm": 9.9202241897583,
+      "learning_rate": 3.966966966966968e-06,
+      "loss": 0.3521,
+      "step": 60270
+    },
+    {
+      "epoch": 181.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5259208679199219,
+      "eval_runtime": 12.9099,
+      "eval_samples_per_second": 774.599,
+      "eval_steps_per_second": 3.098,
+      "step": 60273
+    },
+    {
+      "epoch": 181.02,
+      "grad_norm": 15.941812515258789,
+      "learning_rate": 3.9659659659659664e-06,
+      "loss": 0.4123,
+      "step": 60280
+    },
+    {
+      "epoch": 181.05,
+      "grad_norm": 14.975605010986328,
+      "learning_rate": 3.964964964964965e-06,
+      "loss": 0.3311,
+      "step": 60290
+    },
+    {
+      "epoch": 181.08,
+      "grad_norm": 13.285609245300293,
+      "learning_rate": 3.9639639639639645e-06,
+      "loss": 0.3467,
+      "step": 60300
+    },
+    {
+      "epoch": 181.11,
+      "grad_norm": 18.58491325378418,
+      "learning_rate": 3.962962962962963e-06,
+      "loss": 0.3766,
+      "step": 60310
+    },
+    {
+      "epoch": 181.14,
+      "grad_norm": 12.795409202575684,
+      "learning_rate": 3.961961961961962e-06,
+      "loss": 0.3476,
+      "step": 60320
+    },
+    {
+      "epoch": 181.17,
+      "grad_norm": 21.63353157043457,
+      "learning_rate": 3.960960960960961e-06,
+      "loss": 0.3632,
+      "step": 60330
+    },
+    {
+      "epoch": 181.2,
+      "grad_norm": 16.743820190429688,
+      "learning_rate": 3.9599599599599606e-06,
+      "loss": 0.376,
+      "step": 60340
+    },
+    {
+      "epoch": 181.23,
+      "grad_norm": 30.922998428344727,
+      "learning_rate": 3.958958958958959e-06,
+      "loss": 0.3942,
+      "step": 60350
+    },
+    {
+      "epoch": 181.26,
+      "grad_norm": 8.327676773071289,
+      "learning_rate": 3.957957957957958e-06,
+      "loss": 0.3732,
+      "step": 60360
+    },
+    {
+      "epoch": 181.29,
+      "grad_norm": 13.994136810302734,
+      "learning_rate": 3.956956956956957e-06,
+      "loss": 0.3827,
+      "step": 60370
+    },
+    {
+      "epoch": 181.32,
+      "grad_norm": 17.431140899658203,
+      "learning_rate": 3.955955955955957e-06,
+      "loss": 0.3852,
+      "step": 60380
+    },
+    {
+      "epoch": 181.35,
+      "grad_norm": 18.052934646606445,
+      "learning_rate": 3.954954954954955e-06,
+      "loss": 0.4123,
+      "step": 60390
+    },
+    {
+      "epoch": 181.38,
+      "grad_norm": 12.068872451782227,
+      "learning_rate": 3.953953953953955e-06,
+      "loss": 0.398,
+      "step": 60400
+    },
+    {
+      "epoch": 181.41,
+      "grad_norm": 16.060640335083008,
+      "learning_rate": 3.952952952952953e-06,
+      "loss": 0.363,
+      "step": 60410
+    },
+    {
+      "epoch": 181.44,
+      "grad_norm": 17.949800491333008,
+      "learning_rate": 3.951951951951952e-06,
+      "loss": 0.387,
+      "step": 60420
+    },
+    {
+      "epoch": 181.47,
+      "grad_norm": 16.006328582763672,
+      "learning_rate": 3.950950950950951e-06,
+      "loss": 0.3521,
+      "step": 60430
+    },
+    {
+      "epoch": 181.5,
+      "grad_norm": 14.964012145996094,
+      "learning_rate": 3.949949949949951e-06,
+      "loss": 0.3593,
+      "step": 60440
+    },
+    {
+      "epoch": 181.53,
+      "grad_norm": 17.45253562927246,
+      "learning_rate": 3.948948948948949e-06,
+      "loss": 0.3555,
+      "step": 60450
+    },
+    {
+      "epoch": 181.56,
+      "grad_norm": 18.546960830688477,
+      "learning_rate": 3.947947947947948e-06,
+      "loss": 0.3837,
+      "step": 60460
+    },
+    {
+      "epoch": 181.59,
+      "grad_norm": 17.695154190063477,
+      "learning_rate": 3.946946946946947e-06,
+      "loss": 0.3892,
+      "step": 60470
+    },
+    {
+      "epoch": 181.62,
+      "grad_norm": 17.75796890258789,
+      "learning_rate": 3.945945945945947e-06,
+      "loss": 0.3416,
+      "step": 60480
+    },
+    {
+      "epoch": 181.65,
+      "grad_norm": 18.541603088378906,
+      "learning_rate": 3.944944944944945e-06,
+      "loss": 0.358,
+      "step": 60490
+    },
+    {
+      "epoch": 181.68,
+      "grad_norm": 19.17035675048828,
+      "learning_rate": 3.943943943943944e-06,
+      "loss": 0.3481,
+      "step": 60500
+    },
+    {
+      "epoch": 181.71,
+      "grad_norm": 16.0799617767334,
+      "learning_rate": 3.9429429429429434e-06,
+      "loss": 0.3603,
+      "step": 60510
+    },
+    {
+      "epoch": 181.74,
+      "grad_norm": 14.032349586486816,
+      "learning_rate": 3.941941941941943e-06,
+      "loss": 0.3361,
+      "step": 60520
+    },
+    {
+      "epoch": 181.77,
+      "grad_norm": 17.484325408935547,
+      "learning_rate": 3.940940940940941e-06,
+      "loss": 0.3507,
+      "step": 60530
+    },
+    {
+      "epoch": 181.8,
+      "grad_norm": 13.531590461730957,
+      "learning_rate": 3.93993993993994e-06,
+      "loss": 0.3891,
+      "step": 60540
+    },
+    {
+      "epoch": 181.83,
+      "grad_norm": 13.572278022766113,
+      "learning_rate": 3.9389389389389395e-06,
+      "loss": 0.3456,
+      "step": 60550
+    },
+    {
+      "epoch": 181.86,
+      "grad_norm": 12.967151641845703,
+      "learning_rate": 3.937937937937938e-06,
+      "loss": 0.3522,
+      "step": 60560
+    },
+    {
+      "epoch": 181.89,
+      "grad_norm": 12.550870895385742,
+      "learning_rate": 3.936936936936937e-06,
+      "loss": 0.3456,
+      "step": 60570
+    },
+    {
+      "epoch": 181.92,
+      "grad_norm": 18.572710037231445,
+      "learning_rate": 3.935935935935936e-06,
+      "loss": 0.4076,
+      "step": 60580
+    },
+    {
+      "epoch": 181.95,
+      "grad_norm": 18.53325080871582,
+      "learning_rate": 3.934934934934936e-06,
+      "loss": 0.3864,
+      "step": 60590
+    },
+    {
+      "epoch": 181.98,
+      "grad_norm": 27.028919219970703,
+      "learning_rate": 3.933933933933934e-06,
+      "loss": 0.3922,
+      "step": 60600
+    },
+    {
+      "epoch": 182.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5316426753997803,
+      "eval_runtime": 12.9179,
+      "eval_samples_per_second": 774.12,
+      "eval_steps_per_second": 3.096,
+      "step": 60606
+    },
+    {
+      "epoch": 182.01,
+      "grad_norm": 20.64849853515625,
+      "learning_rate": 3.932932932932933e-06,
+      "loss": 0.4385,
+      "step": 60610
+    },
+    {
+      "epoch": 182.04,
+      "grad_norm": 20.17979621887207,
+      "learning_rate": 3.931931931931932e-06,
+      "loss": 0.3892,
+      "step": 60620
+    },
+    {
+      "epoch": 182.07,
+      "grad_norm": 17.239627838134766,
+      "learning_rate": 3.930930930930932e-06,
+      "loss": 0.3483,
+      "step": 60630
+    },
+    {
+      "epoch": 182.1,
+      "grad_norm": 17.36978530883789,
+      "learning_rate": 3.92992992992993e-06,
+      "loss": 0.4094,
+      "step": 60640
+    },
+    {
+      "epoch": 182.13,
+      "grad_norm": 17.123775482177734,
+      "learning_rate": 3.92892892892893e-06,
+      "loss": 0.3795,
+      "step": 60650
+    },
+    {
+      "epoch": 182.16,
+      "grad_norm": 22.18364143371582,
+      "learning_rate": 3.927927927927928e-06,
+      "loss": 0.41,
+      "step": 60660
+    },
+    {
+      "epoch": 182.19,
+      "grad_norm": 14.032144546508789,
+      "learning_rate": 3.926926926926927e-06,
+      "loss": 0.3374,
+      "step": 60670
+    },
+    {
+      "epoch": 182.22,
+      "grad_norm": 14.208514213562012,
+      "learning_rate": 3.925925925925926e-06,
+      "loss": 0.421,
+      "step": 60680
+    },
+    {
+      "epoch": 182.25,
+      "grad_norm": 12.743430137634277,
+      "learning_rate": 3.924924924924926e-06,
+      "loss": 0.3855,
+      "step": 60690
+    },
+    {
+      "epoch": 182.28,
+      "grad_norm": 25.91026496887207,
+      "learning_rate": 3.923923923923924e-06,
+      "loss": 0.3246,
+      "step": 60700
+    },
+    {
+      "epoch": 182.31,
+      "grad_norm": 11.527327537536621,
+      "learning_rate": 3.922922922922923e-06,
+      "loss": 0.3971,
+      "step": 60710
+    },
+    {
+      "epoch": 182.34,
+      "grad_norm": 20.978164672851562,
+      "learning_rate": 3.921921921921922e-06,
+      "loss": 0.3895,
+      "step": 60720
+    },
+    {
+      "epoch": 182.37,
+      "grad_norm": 11.694549560546875,
+      "learning_rate": 3.920920920920922e-06,
+      "loss": 0.3249,
+      "step": 60730
+    },
+    {
+      "epoch": 182.4,
+      "grad_norm": 28.196826934814453,
+      "learning_rate": 3.9199199199199204e-06,
+      "loss": 0.3331,
+      "step": 60740
+    },
+    {
+      "epoch": 182.43,
+      "grad_norm": 13.628348350524902,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.4082,
+      "step": 60750
+    },
+    {
+      "epoch": 182.46,
+      "grad_norm": 17.187551498413086,
+      "learning_rate": 3.9179179179179185e-06,
+      "loss": 0.4386,
+      "step": 60760
+    },
+    {
+      "epoch": 182.49,
+      "grad_norm": 14.947317123413086,
+      "learning_rate": 3.916916916916917e-06,
+      "loss": 0.3753,
+      "step": 60770
+    },
+    {
+      "epoch": 182.52,
+      "grad_norm": 15.905251502990723,
+      "learning_rate": 3.915915915915916e-06,
+      "loss": 0.3348,
+      "step": 60780
+    },
+    {
+      "epoch": 182.55,
+      "grad_norm": 23.328073501586914,
+      "learning_rate": 3.914914914914915e-06,
+      "loss": 0.3764,
+      "step": 60790
+    },
+    {
+      "epoch": 182.58,
+      "grad_norm": 16.816490173339844,
+      "learning_rate": 3.9139139139139145e-06,
+      "loss": 0.3455,
+      "step": 60800
+    },
+    {
+      "epoch": 182.61,
+      "grad_norm": 21.183475494384766,
+      "learning_rate": 3.912912912912913e-06,
+      "loss": 0.3787,
+      "step": 60810
+    },
+    {
+      "epoch": 182.64,
+      "grad_norm": 13.397309303283691,
+      "learning_rate": 3.911911911911912e-06,
+      "loss": 0.3403,
+      "step": 60820
+    },
+    {
+      "epoch": 182.67,
+      "grad_norm": 15.193707466125488,
+      "learning_rate": 3.910910910910911e-06,
+      "loss": 0.4224,
+      "step": 60830
+    },
+    {
+      "epoch": 182.7,
+      "grad_norm": 20.38916778564453,
+      "learning_rate": 3.909909909909911e-06,
+      "loss": 0.4185,
+      "step": 60840
+    },
+    {
+      "epoch": 182.73,
+      "grad_norm": 13.88211441040039,
+      "learning_rate": 3.908908908908909e-06,
+      "loss": 0.3283,
+      "step": 60850
+    },
+    {
+      "epoch": 182.76,
+      "grad_norm": 12.186492919921875,
+      "learning_rate": 3.907907907907908e-06,
+      "loss": 0.3925,
+      "step": 60860
+    },
+    {
+      "epoch": 182.79,
+      "grad_norm": 21.197725296020508,
+      "learning_rate": 3.906906906906907e-06,
+      "loss": 0.3838,
+      "step": 60870
+    },
+    {
+      "epoch": 182.82,
+      "grad_norm": 12.530217170715332,
+      "learning_rate": 3.905905905905906e-06,
+      "loss": 0.4084,
+      "step": 60880
+    },
+    {
+      "epoch": 182.85,
+      "grad_norm": 12.80907917022705,
+      "learning_rate": 3.904904904904905e-06,
+      "loss": 0.3358,
+      "step": 60890
+    },
+    {
+      "epoch": 182.88,
+      "grad_norm": 15.354159355163574,
+      "learning_rate": 3.903903903903904e-06,
+      "loss": 0.3584,
+      "step": 60900
+    },
+    {
+      "epoch": 182.91,
+      "grad_norm": 21.151884078979492,
+      "learning_rate": 3.902902902902903e-06,
+      "loss": 0.3754,
+      "step": 60910
+    },
+    {
+      "epoch": 182.94,
+      "grad_norm": 14.971238136291504,
+      "learning_rate": 3.901901901901902e-06,
+      "loss": 0.368,
+      "step": 60920
+    },
+    {
+      "epoch": 182.97,
+      "grad_norm": 19.604110717773438,
+      "learning_rate": 3.900900900900901e-06,
+      "loss": 0.3391,
+      "step": 60930
+    },
+    {
+      "epoch": 183.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5316169261932373,
+      "eval_runtime": 12.649,
+      "eval_samples_per_second": 790.578,
+      "eval_steps_per_second": 3.162,
+      "step": 60939
+    },
+    {
+      "epoch": 183.0,
+      "grad_norm": 18.611833572387695,
+      "learning_rate": 3.899899899899901e-06,
+      "loss": 0.3479,
+      "step": 60940
+    },
+    {
+      "epoch": 183.03,
+      "grad_norm": 12.535316467285156,
+      "learning_rate": 3.898898898898899e-06,
+      "loss": 0.3492,
+      "step": 60950
+    },
+    {
+      "epoch": 183.06,
+      "grad_norm": 23.636775970458984,
+      "learning_rate": 3.897897897897898e-06,
+      "loss": 0.3511,
+      "step": 60960
+    },
+    {
+      "epoch": 183.09,
+      "grad_norm": 15.51108169555664,
+      "learning_rate": 3.8968968968968974e-06,
+      "loss": 0.3681,
+      "step": 60970
+    },
+    {
+      "epoch": 183.12,
+      "grad_norm": 16.654783248901367,
+      "learning_rate": 3.895895895895896e-06,
+      "loss": 0.3399,
+      "step": 60980
+    },
+    {
+      "epoch": 183.15,
+      "grad_norm": 14.904403686523438,
+      "learning_rate": 3.894894894894895e-06,
+      "loss": 0.3926,
+      "step": 60990
+    },
+    {
+      "epoch": 183.18,
+      "grad_norm": 25.373088836669922,
+      "learning_rate": 3.893893893893894e-06,
+      "loss": 0.3494,
+      "step": 61000
+    },
+    {
+      "epoch": 183.21,
+      "grad_norm": 27.35822868347168,
+      "learning_rate": 3.8928928928928935e-06,
+      "loss": 0.3978,
+      "step": 61010
+    },
+    {
+      "epoch": 183.24,
+      "grad_norm": 14.119414329528809,
+      "learning_rate": 3.891891891891892e-06,
+      "loss": 0.3631,
+      "step": 61020
+    },
+    {
+      "epoch": 183.27,
+      "grad_norm": 17.848554611206055,
+      "learning_rate": 3.890890890890891e-06,
+      "loss": 0.368,
+      "step": 61030
+    },
+    {
+      "epoch": 183.3,
+      "grad_norm": 13.963902473449707,
+      "learning_rate": 3.88988988988989e-06,
+      "loss": 0.3604,
+      "step": 61040
+    },
+    {
+      "epoch": 183.33,
+      "grad_norm": 33.026954650878906,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.3279,
+      "step": 61050
+    },
+    {
+      "epoch": 183.36,
+      "grad_norm": 10.560956001281738,
+      "learning_rate": 3.887887887887888e-06,
+      "loss": 0.3499,
+      "step": 61060
+    },
+    {
+      "epoch": 183.39,
+      "grad_norm": 16.34345054626465,
+      "learning_rate": 3.886886886886887e-06,
+      "loss": 0.3498,
+      "step": 61070
+    },
+    {
+      "epoch": 183.42,
+      "grad_norm": 19.206296920776367,
+      "learning_rate": 3.885885885885886e-06,
+      "loss": 0.355,
+      "step": 61080
+    },
+    {
+      "epoch": 183.45,
+      "grad_norm": 9.520086288452148,
+      "learning_rate": 3.884884884884886e-06,
+      "loss": 0.3859,
+      "step": 61090
+    },
+    {
+      "epoch": 183.48,
+      "grad_norm": 12.109062194824219,
+      "learning_rate": 3.883883883883884e-06,
+      "loss": 0.3886,
+      "step": 61100
+    },
+    {
+      "epoch": 183.51,
+      "grad_norm": 13.966768264770508,
+      "learning_rate": 3.882882882882883e-06,
+      "loss": 0.3242,
+      "step": 61110
+    },
+    {
+      "epoch": 183.54,
+      "grad_norm": 25.05291748046875,
+      "learning_rate": 3.881881881881882e-06,
+      "loss": 0.3501,
+      "step": 61120
+    },
+    {
+      "epoch": 183.57,
+      "grad_norm": 13.17016315460205,
+      "learning_rate": 3.880880880880881e-06,
+      "loss": 0.3664,
+      "step": 61130
+    },
+    {
+      "epoch": 183.6,
+      "grad_norm": 20.776029586791992,
+      "learning_rate": 3.87987987987988e-06,
+      "loss": 0.3699,
+      "step": 61140
+    },
+    {
+      "epoch": 183.63,
+      "grad_norm": 9.171614646911621,
+      "learning_rate": 3.878878878878879e-06,
+      "loss": 0.3352,
+      "step": 61150
+    },
+    {
+      "epoch": 183.66,
+      "grad_norm": 12.49626350402832,
+      "learning_rate": 3.877877877877878e-06,
+      "loss": 0.3647,
+      "step": 61160
+    },
+    {
+      "epoch": 183.69,
+      "grad_norm": 33.79391860961914,
+      "learning_rate": 3.876876876876877e-06,
+      "loss": 0.3921,
+      "step": 61170
+    },
+    {
+      "epoch": 183.72,
+      "grad_norm": 14.179243087768555,
+      "learning_rate": 3.875875875875876e-06,
+      "loss": 0.3953,
+      "step": 61180
+    },
+    {
+      "epoch": 183.75,
+      "grad_norm": 15.619384765625,
+      "learning_rate": 3.874874874874876e-06,
+      "loss": 0.3286,
+      "step": 61190
+    },
+    {
+      "epoch": 183.78,
+      "grad_norm": 17.025611877441406,
+      "learning_rate": 3.8738738738738744e-06,
+      "loss": 0.3763,
+      "step": 61200
+    },
+    {
+      "epoch": 183.81,
+      "grad_norm": 23.923660278320312,
+      "learning_rate": 3.872872872872873e-06,
+      "loss": 0.3797,
+      "step": 61210
+    },
+    {
+      "epoch": 183.84,
+      "grad_norm": 24.075496673583984,
+      "learning_rate": 3.8718718718718725e-06,
+      "loss": 0.3986,
+      "step": 61220
+    },
+    {
+      "epoch": 183.87,
+      "grad_norm": 19.645830154418945,
+      "learning_rate": 3.870870870870871e-06,
+      "loss": 0.3945,
+      "step": 61230
+    },
+    {
+      "epoch": 183.9,
+      "grad_norm": 14.014632225036621,
+      "learning_rate": 3.86986986986987e-06,
+      "loss": 0.3116,
+      "step": 61240
+    },
+    {
+      "epoch": 183.93,
+      "grad_norm": 19.020322799682617,
+      "learning_rate": 3.868868868868869e-06,
+      "loss": 0.373,
+      "step": 61250
+    },
+    {
+      "epoch": 183.96,
+      "grad_norm": 20.4427547454834,
+      "learning_rate": 3.8678678678678685e-06,
+      "loss": 0.4435,
+      "step": 61260
+    },
+    {
+      "epoch": 183.99,
+      "grad_norm": 16.334232330322266,
+      "learning_rate": 3.866866866866867e-06,
+      "loss": 0.3646,
+      "step": 61270
+    },
+    {
+      "epoch": 184.0,
+      "eval_accuracy": 0.8637,
+      "eval_loss": 0.5329204797744751,
+      "eval_runtime": 12.7819,
+      "eval_samples_per_second": 782.353,
+      "eval_steps_per_second": 3.129,
+      "step": 61272
+    },
+    {
+      "epoch": 184.02,
+      "grad_norm": 9.615008354187012,
+      "learning_rate": 3.865865865865866e-06,
+      "loss": 0.2948,
+      "step": 61280
+    },
+    {
+      "epoch": 184.05,
+      "grad_norm": 23.21010398864746,
+      "learning_rate": 3.864864864864865e-06,
+      "loss": 0.3754,
+      "step": 61290
+    },
+    {
+      "epoch": 184.08,
+      "grad_norm": 18.36243438720703,
+      "learning_rate": 3.863863863863865e-06,
+      "loss": 0.303,
+      "step": 61300
+    },
+    {
+      "epoch": 184.11,
+      "grad_norm": 12.67408561706543,
+      "learning_rate": 3.862862862862863e-06,
+      "loss": 0.3633,
+      "step": 61310
+    },
+    {
+      "epoch": 184.14,
+      "grad_norm": 17.231964111328125,
+      "learning_rate": 3.861861861861862e-06,
+      "loss": 0.389,
+      "step": 61320
+    },
+    {
+      "epoch": 184.17,
+      "grad_norm": 16.66814613342285,
+      "learning_rate": 3.860860860860861e-06,
+      "loss": 0.3874,
+      "step": 61330
+    },
+    {
+      "epoch": 184.2,
+      "grad_norm": 24.904090881347656,
+      "learning_rate": 3.85985985985986e-06,
+      "loss": 0.3825,
+      "step": 61340
+    },
+    {
+      "epoch": 184.23,
+      "grad_norm": 22.55470085144043,
+      "learning_rate": 3.858858858858859e-06,
+      "loss": 0.3867,
+      "step": 61350
+    },
+    {
+      "epoch": 184.26,
+      "grad_norm": 17.097288131713867,
+      "learning_rate": 3.857857857857858e-06,
+      "loss": 0.3265,
+      "step": 61360
+    },
+    {
+      "epoch": 184.29,
+      "grad_norm": 13.741023063659668,
+      "learning_rate": 3.856856856856857e-06,
+      "loss": 0.3347,
+      "step": 61370
+    },
+    {
+      "epoch": 184.32,
+      "grad_norm": 18.142595291137695,
+      "learning_rate": 3.855855855855856e-06,
+      "loss": 0.3694,
+      "step": 61380
+    },
+    {
+      "epoch": 184.35,
+      "grad_norm": 11.918787956237793,
+      "learning_rate": 3.854854854854855e-06,
+      "loss": 0.3981,
+      "step": 61390
+    },
+    {
+      "epoch": 184.38,
+      "grad_norm": 15.91456413269043,
+      "learning_rate": 3.853853853853854e-06,
+      "loss": 0.3293,
+      "step": 61400
+    },
+    {
+      "epoch": 184.41,
+      "grad_norm": 21.000581741333008,
+      "learning_rate": 3.852852852852853e-06,
+      "loss": 0.4325,
+      "step": 61410
+    },
+    {
+      "epoch": 184.44,
+      "grad_norm": 21.06557273864746,
+      "learning_rate": 3.851851851851852e-06,
+      "loss": 0.3743,
+      "step": 61420
+    },
+    {
+      "epoch": 184.47,
+      "grad_norm": 12.748741149902344,
+      "learning_rate": 3.8508508508508514e-06,
+      "loss": 0.3727,
+      "step": 61430
+    },
+    {
+      "epoch": 184.5,
+      "grad_norm": 27.672882080078125,
+      "learning_rate": 3.84984984984985e-06,
+      "loss": 0.3868,
+      "step": 61440
+    },
+    {
+      "epoch": 184.53,
+      "grad_norm": 16.715557098388672,
+      "learning_rate": 3.848848848848849e-06,
+      "loss": 0.4138,
+      "step": 61450
+    },
+    {
+      "epoch": 184.56,
+      "grad_norm": 16.810789108276367,
+      "learning_rate": 3.847847847847848e-06,
+      "loss": 0.385,
+      "step": 61460
+    },
+    {
+      "epoch": 184.59,
+      "grad_norm": 24.70064926147461,
+      "learning_rate": 3.8468468468468475e-06,
+      "loss": 0.3739,
+      "step": 61470
+    },
+    {
+      "epoch": 184.62,
+      "grad_norm": 10.757923126220703,
+      "learning_rate": 3.845845845845846e-06,
+      "loss": 0.3411,
+      "step": 61480
+    },
+    {
+      "epoch": 184.65,
+      "grad_norm": 17.468215942382812,
+      "learning_rate": 3.844844844844845e-06,
+      "loss": 0.3606,
+      "step": 61490
+    },
+    {
+      "epoch": 184.68,
+      "grad_norm": 13.585052490234375,
+      "learning_rate": 3.843843843843844e-06,
+      "loss": 0.3477,
+      "step": 61500
+    },
+    {
+      "epoch": 184.71,
+      "grad_norm": 14.551653861999512,
+      "learning_rate": 3.842842842842844e-06,
+      "loss": 0.3276,
+      "step": 61510
+    },
+    {
+      "epoch": 184.74,
+      "grad_norm": 21.039020538330078,
+      "learning_rate": 3.841841841841842e-06,
+      "loss": 0.366,
+      "step": 61520
+    },
+    {
+      "epoch": 184.77,
+      "grad_norm": 25.306793212890625,
+      "learning_rate": 3.840840840840841e-06,
+      "loss": 0.3747,
+      "step": 61530
+    },
+    {
+      "epoch": 184.8,
+      "grad_norm": 10.584951400756836,
+      "learning_rate": 3.83983983983984e-06,
+      "loss": 0.3515,
+      "step": 61540
+    },
+    {
+      "epoch": 184.83,
+      "grad_norm": 12.91821575164795,
+      "learning_rate": 3.83883883883884e-06,
+      "loss": 0.3296,
+      "step": 61550
+    },
+    {
+      "epoch": 184.86,
+      "grad_norm": 13.807701110839844,
+      "learning_rate": 3.837837837837838e-06,
+      "loss": 0.3756,
+      "step": 61560
+    },
+    {
+      "epoch": 184.89,
+      "grad_norm": 30.14850616455078,
+      "learning_rate": 3.836836836836837e-06,
+      "loss": 0.3914,
+      "step": 61570
+    },
+    {
+      "epoch": 184.92,
+      "grad_norm": 15.080806732177734,
+      "learning_rate": 3.835835835835836e-06,
+      "loss": 0.39,
+      "step": 61580
+    },
+    {
+      "epoch": 184.95,
+      "grad_norm": 16.957616806030273,
+      "learning_rate": 3.834834834834835e-06,
+      "loss": 0.351,
+      "step": 61590
+    },
+    {
+      "epoch": 184.98,
+      "grad_norm": 20.09813690185547,
+      "learning_rate": 3.833833833833834e-06,
+      "loss": 0.4033,
+      "step": 61600
+    },
+    {
+      "epoch": 185.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.5357367396354675,
+      "eval_runtime": 12.7258,
+      "eval_samples_per_second": 785.807,
+      "eval_steps_per_second": 3.143,
+      "step": 61605
+    },
+    {
+      "epoch": 185.02,
+      "grad_norm": 20.655670166015625,
+      "learning_rate": 3.832832832832833e-06,
+      "loss": 0.3586,
+      "step": 61610
+    },
+    {
+      "epoch": 185.05,
+      "grad_norm": 11.373549461364746,
+      "learning_rate": 3.831831831831832e-06,
+      "loss": 0.3706,
+      "step": 61620
+    },
+    {
+      "epoch": 185.08,
+      "grad_norm": 13.904773712158203,
+      "learning_rate": 3.830830830830831e-06,
+      "loss": 0.3812,
+      "step": 61630
+    },
+    {
+      "epoch": 185.11,
+      "grad_norm": 21.287883758544922,
+      "learning_rate": 3.82982982982983e-06,
+      "loss": 0.3338,
+      "step": 61640
+    },
+    {
+      "epoch": 185.14,
+      "grad_norm": 29.931888580322266,
+      "learning_rate": 3.828828828828829e-06,
+      "loss": 0.3491,
+      "step": 61650
+    },
+    {
+      "epoch": 185.17,
+      "grad_norm": 18.118667602539062,
+      "learning_rate": 3.8278278278278284e-06,
+      "loss": 0.3392,
+      "step": 61660
+    },
+    {
+      "epoch": 185.2,
+      "grad_norm": 15.445931434631348,
+      "learning_rate": 3.826826826826827e-06,
+      "loss": 0.3345,
+      "step": 61670
+    },
+    {
+      "epoch": 185.23,
+      "grad_norm": 27.475629806518555,
+      "learning_rate": 3.8258258258258265e-06,
+      "loss": 0.387,
+      "step": 61680
+    },
+    {
+      "epoch": 185.26,
+      "grad_norm": 25.907787322998047,
+      "learning_rate": 3.824824824824825e-06,
+      "loss": 0.3709,
+      "step": 61690
+    },
+    {
+      "epoch": 185.29,
+      "grad_norm": 16.811561584472656,
+      "learning_rate": 3.823823823823824e-06,
+      "loss": 0.4095,
+      "step": 61700
+    },
+    {
+      "epoch": 185.32,
+      "grad_norm": 14.155433654785156,
+      "learning_rate": 3.822822822822823e-06,
+      "loss": 0.3564,
+      "step": 61710
+    },
+    {
+      "epoch": 185.35,
+      "grad_norm": 13.723014831542969,
+      "learning_rate": 3.8218218218218225e-06,
+      "loss": 0.3962,
+      "step": 61720
+    },
+    {
+      "epoch": 185.38,
+      "grad_norm": 18.712215423583984,
+      "learning_rate": 3.820820820820821e-06,
+      "loss": 0.383,
+      "step": 61730
+    },
+    {
+      "epoch": 185.41,
+      "grad_norm": 18.567123413085938,
+      "learning_rate": 3.81981981981982e-06,
+      "loss": 0.3702,
+      "step": 61740
+    },
+    {
+      "epoch": 185.44,
+      "grad_norm": 24.38343048095703,
+      "learning_rate": 3.818818818818819e-06,
+      "loss": 0.3554,
+      "step": 61750
+    },
+    {
+      "epoch": 185.47,
+      "grad_norm": 15.07786750793457,
+      "learning_rate": 3.817817817817819e-06,
+      "loss": 0.3792,
+      "step": 61760
+    },
+    {
+      "epoch": 185.5,
+      "grad_norm": 14.772867202758789,
+      "learning_rate": 3.816816816816817e-06,
+      "loss": 0.4069,
+      "step": 61770
+    },
+    {
+      "epoch": 185.53,
+      "grad_norm": 19.435298919677734,
+      "learning_rate": 3.815815815815816e-06,
+      "loss": 0.3494,
+      "step": 61780
+    },
+    {
+      "epoch": 185.56,
+      "grad_norm": 9.065760612487793,
+      "learning_rate": 3.814814814814815e-06,
+      "loss": 0.3707,
+      "step": 61790
+    },
+    {
+      "epoch": 185.59,
+      "grad_norm": 17.157493591308594,
+      "learning_rate": 3.8138138138138143e-06,
+      "loss": 0.3367,
+      "step": 61800
+    },
+    {
+      "epoch": 185.62,
+      "grad_norm": 13.699775695800781,
+      "learning_rate": 3.8128128128128133e-06,
+      "loss": 0.3259,
+      "step": 61810
+    },
+    {
+      "epoch": 185.65,
+      "grad_norm": 31.067916870117188,
+      "learning_rate": 3.811811811811812e-06,
+      "loss": 0.4052,
+      "step": 61820
+    },
+    {
+      "epoch": 185.68,
+      "grad_norm": 13.346184730529785,
+      "learning_rate": 3.810810810810811e-06,
+      "loss": 0.2861,
+      "step": 61830
+    },
+    {
+      "epoch": 185.71,
+      "grad_norm": 15.237887382507324,
+      "learning_rate": 3.8098098098098103e-06,
+      "loss": 0.4054,
+      "step": 61840
+    },
+    {
+      "epoch": 185.74,
+      "grad_norm": 14.167617797851562,
+      "learning_rate": 3.8088088088088094e-06,
+      "loss": 0.345,
+      "step": 61850
+    },
+    {
+      "epoch": 185.77,
+      "grad_norm": 14.548701286315918,
+      "learning_rate": 3.807807807807808e-06,
+      "loss": 0.3458,
+      "step": 61860
+    },
+    {
+      "epoch": 185.8,
+      "grad_norm": 17.270980834960938,
+      "learning_rate": 3.806806806806807e-06,
+      "loss": 0.3476,
+      "step": 61870
+    },
+    {
+      "epoch": 185.83,
+      "grad_norm": 17.77896499633789,
+      "learning_rate": 3.805805805805806e-06,
+      "loss": 0.3956,
+      "step": 61880
+    },
+    {
+      "epoch": 185.86,
+      "grad_norm": 12.438741683959961,
+      "learning_rate": 3.8048048048048054e-06,
+      "loss": 0.3412,
+      "step": 61890
+    },
+    {
+      "epoch": 185.89,
+      "grad_norm": 16.009599685668945,
+      "learning_rate": 3.803803803803804e-06,
+      "loss": 0.3686,
+      "step": 61900
+    },
+    {
+      "epoch": 185.92,
+      "grad_norm": 11.632037162780762,
+      "learning_rate": 3.802802802802803e-06,
+      "loss": 0.3369,
+      "step": 61910
+    },
+    {
+      "epoch": 185.95,
+      "grad_norm": 17.461368560791016,
+      "learning_rate": 3.801801801801802e-06,
+      "loss": 0.3571,
+      "step": 61920
+    },
+    {
+      "epoch": 185.98,
+      "grad_norm": 21.06696319580078,
+      "learning_rate": 3.800800800800801e-06,
+      "loss": 0.395,
+      "step": 61930
+    },
+    {
+      "epoch": 186.0,
+      "eval_accuracy": 0.8634,
+      "eval_loss": 0.5375855565071106,
+      "eval_runtime": 12.5697,
+      "eval_samples_per_second": 795.562,
+      "eval_steps_per_second": 3.182,
+      "step": 61938
+    },
+    {
+      "epoch": 186.01,
+      "grad_norm": 18.78667640686035,
+      "learning_rate": 3.7997997997997997e-06,
+      "loss": 0.3394,
+      "step": 61940
+    },
+    {
+      "epoch": 186.04,
+      "grad_norm": 12.89063549041748,
+      "learning_rate": 3.798798798798799e-06,
+      "loss": 0.3655,
+      "step": 61950
+    },
+    {
+      "epoch": 186.07,
+      "grad_norm": 18.537227630615234,
+      "learning_rate": 3.797797797797798e-06,
+      "loss": 0.376,
+      "step": 61960
+    },
+    {
+      "epoch": 186.1,
+      "grad_norm": 11.462427139282227,
+      "learning_rate": 3.796796796796797e-06,
+      "loss": 0.3555,
+      "step": 61970
+    },
+    {
+      "epoch": 186.13,
+      "grad_norm": 12.742891311645508,
+      "learning_rate": 3.795795795795796e-06,
+      "loss": 0.3541,
+      "step": 61980
+    },
+    {
+      "epoch": 186.16,
+      "grad_norm": 12.687728881835938,
+      "learning_rate": 3.7947947947947948e-06,
+      "loss": 0.3785,
+      "step": 61990
+    },
+    {
+      "epoch": 186.19,
+      "grad_norm": 17.527986526489258,
+      "learning_rate": 3.793793793793794e-06,
+      "loss": 0.3735,
+      "step": 62000
+    },
+    {
+      "epoch": 186.22,
+      "grad_norm": 19.727144241333008,
+      "learning_rate": 3.7927927927927932e-06,
+      "loss": 0.4365,
+      "step": 62010
+    },
+    {
+      "epoch": 186.25,
+      "grad_norm": 15.677107810974121,
+      "learning_rate": 3.7917917917917922e-06,
+      "loss": 0.3697,
+      "step": 62020
+    },
+    {
+      "epoch": 186.28,
+      "grad_norm": 13.216383934020996,
+      "learning_rate": 3.790790790790791e-06,
+      "loss": 0.4098,
+      "step": 62030
+    },
+    {
+      "epoch": 186.31,
+      "grad_norm": 17.502975463867188,
+      "learning_rate": 3.78978978978979e-06,
+      "loss": 0.409,
+      "step": 62040
+    },
+    {
+      "epoch": 186.34,
+      "grad_norm": 25.838735580444336,
+      "learning_rate": 3.7887887887887893e-06,
+      "loss": 0.3433,
+      "step": 62050
+    },
+    {
+      "epoch": 186.37,
+      "grad_norm": 21.889413833618164,
+      "learning_rate": 3.7877877877877883e-06,
+      "loss": 0.338,
+      "step": 62060
+    },
+    {
+      "epoch": 186.4,
+      "grad_norm": 12.717191696166992,
+      "learning_rate": 3.786786786786787e-06,
+      "loss": 0.3598,
+      "step": 62070
+    },
+    {
+      "epoch": 186.43,
+      "grad_norm": 11.254862785339355,
+      "learning_rate": 3.785785785785786e-06,
+      "loss": 0.3552,
+      "step": 62080
+    },
+    {
+      "epoch": 186.46,
+      "grad_norm": 12.973278999328613,
+      "learning_rate": 3.7847847847847854e-06,
+      "loss": 0.3833,
+      "step": 62090
+    },
+    {
+      "epoch": 186.49,
+      "grad_norm": 15.253023147583008,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.3837,
+      "step": 62100
+    },
+    {
+      "epoch": 186.52,
+      "grad_norm": 23.961027145385742,
+      "learning_rate": 3.782782782782783e-06,
+      "loss": 0.3972,
+      "step": 62110
+    },
+    {
+      "epoch": 186.55,
+      "grad_norm": 16.240331649780273,
+      "learning_rate": 3.781781781781782e-06,
+      "loss": 0.3228,
+      "step": 62120
+    },
+    {
+      "epoch": 186.58,
+      "grad_norm": 23.221376419067383,
+      "learning_rate": 3.780780780780781e-06,
+      "loss": 0.3444,
+      "step": 62130
+    },
+    {
+      "epoch": 186.61,
+      "grad_norm": 11.74186897277832,
+      "learning_rate": 3.7797797797797805e-06,
+      "loss": 0.364,
+      "step": 62140
+    },
+    {
+      "epoch": 186.64,
+      "grad_norm": 15.132006645202637,
+      "learning_rate": 3.778778778778779e-06,
+      "loss": 0.3859,
+      "step": 62150
+    },
+    {
+      "epoch": 186.67,
+      "grad_norm": 11.972746849060059,
+      "learning_rate": 3.777777777777778e-06,
+      "loss": 0.3569,
+      "step": 62160
+    },
+    {
+      "epoch": 186.7,
+      "grad_norm": 9.702228546142578,
+      "learning_rate": 3.776776776776777e-06,
+      "loss": 0.3036,
+      "step": 62170
+    },
+    {
+      "epoch": 186.73,
+      "grad_norm": 18.582801818847656,
+      "learning_rate": 3.775775775775776e-06,
+      "loss": 0.3778,
+      "step": 62180
+    },
+    {
+      "epoch": 186.76,
+      "grad_norm": 21.37168312072754,
+      "learning_rate": 3.7747747747747747e-06,
+      "loss": 0.3694,
+      "step": 62190
+    },
+    {
+      "epoch": 186.79,
+      "grad_norm": 14.785737037658691,
+      "learning_rate": 3.773773773773774e-06,
+      "loss": 0.3867,
+      "step": 62200
+    },
+    {
+      "epoch": 186.82,
+      "grad_norm": 27.627897262573242,
+      "learning_rate": 3.772772772772773e-06,
+      "loss": 0.3706,
+      "step": 62210
+    },
+    {
+      "epoch": 186.85,
+      "grad_norm": 15.21142864227295,
+      "learning_rate": 3.771771771771772e-06,
+      "loss": 0.405,
+      "step": 62220
+    },
+    {
+      "epoch": 186.88,
+      "grad_norm": 23.28750991821289,
+      "learning_rate": 3.7707707707707708e-06,
+      "loss": 0.3723,
+      "step": 62230
+    },
+    {
+      "epoch": 186.91,
+      "grad_norm": 17.11140251159668,
+      "learning_rate": 3.76976976976977e-06,
+      "loss": 0.3397,
+      "step": 62240
+    },
+    {
+      "epoch": 186.94,
+      "grad_norm": 13.164200782775879,
+      "learning_rate": 3.7687687687687692e-06,
+      "loss": 0.3603,
+      "step": 62250
+    },
+    {
+      "epoch": 186.97,
+      "grad_norm": 13.639315605163574,
+      "learning_rate": 3.7677677677677683e-06,
+      "loss": 0.3738,
+      "step": 62260
+    },
+    {
+      "epoch": 187.0,
+      "grad_norm": 14.78702449798584,
+      "learning_rate": 3.7667667667667673e-06,
+      "loss": 0.3253,
+      "step": 62270
+    },
+    {
+      "epoch": 187.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.5345723628997803,
+      "eval_runtime": 12.563,
+      "eval_samples_per_second": 795.989,
+      "eval_steps_per_second": 3.184,
+      "step": 62271
+    },
+    {
+      "epoch": 187.03,
+      "grad_norm": 16.975339889526367,
+      "learning_rate": 3.765765765765766e-06,
+      "loss": 0.3413,
+      "step": 62280
+    },
+    {
+      "epoch": 187.06,
+      "grad_norm": 22.111270904541016,
+      "learning_rate": 3.764764764764765e-06,
+      "loss": 0.3581,
+      "step": 62290
+    },
+    {
+      "epoch": 187.09,
+      "grad_norm": 14.227043151855469,
+      "learning_rate": 3.7637637637637643e-06,
+      "loss": 0.3399,
+      "step": 62300
+    },
+    {
+      "epoch": 187.12,
+      "grad_norm": 24.467906951904297,
+      "learning_rate": 3.7627627627627634e-06,
+      "loss": 0.4331,
+      "step": 62310
+    },
+    {
+      "epoch": 187.15,
+      "grad_norm": 9.9598970413208,
+      "learning_rate": 3.761761761761762e-06,
+      "loss": 0.3634,
+      "step": 62320
+    },
+    {
+      "epoch": 187.18,
+      "grad_norm": 15.418505668640137,
+      "learning_rate": 3.760760760760761e-06,
+      "loss": 0.3791,
+      "step": 62330
+    },
+    {
+      "epoch": 187.21,
+      "grad_norm": 11.169529914855957,
+      "learning_rate": 3.75975975975976e-06,
+      "loss": 0.3087,
+      "step": 62340
+    },
+    {
+      "epoch": 187.24,
+      "grad_norm": 14.740416526794434,
+      "learning_rate": 3.7587587587587594e-06,
+      "loss": 0.3791,
+      "step": 62350
+    },
+    {
+      "epoch": 187.27,
+      "grad_norm": 16.443458557128906,
+      "learning_rate": 3.757757757757758e-06,
+      "loss": 0.2902,
+      "step": 62360
+    },
+    {
+      "epoch": 187.3,
+      "grad_norm": 18.338552474975586,
+      "learning_rate": 3.756756756756757e-06,
+      "loss": 0.3557,
+      "step": 62370
+    },
+    {
+      "epoch": 187.33,
+      "grad_norm": 15.574721336364746,
+      "learning_rate": 3.755755755755756e-06,
+      "loss": 0.371,
+      "step": 62380
+    },
+    {
+      "epoch": 187.36,
+      "grad_norm": 27.96323013305664,
+      "learning_rate": 3.754754754754755e-06,
+      "loss": 0.312,
+      "step": 62390
+    },
+    {
+      "epoch": 187.39,
+      "grad_norm": 23.75743293762207,
+      "learning_rate": 3.7537537537537537e-06,
+      "loss": 0.3864,
+      "step": 62400
+    },
+    {
+      "epoch": 187.42,
+      "grad_norm": 22.34495735168457,
+      "learning_rate": 3.752752752752753e-06,
+      "loss": 0.3829,
+      "step": 62410
+    },
+    {
+      "epoch": 187.45,
+      "grad_norm": 15.213949203491211,
+      "learning_rate": 3.751751751751752e-06,
+      "loss": 0.3775,
+      "step": 62420
+    },
+    {
+      "epoch": 187.48,
+      "grad_norm": 14.607603073120117,
+      "learning_rate": 3.750750750750751e-06,
+      "loss": 0.354,
+      "step": 62430
+    },
+    {
+      "epoch": 187.51,
+      "grad_norm": 16.347990036010742,
+      "learning_rate": 3.7497497497497497e-06,
+      "loss": 0.3515,
+      "step": 62440
+    },
+    {
+      "epoch": 187.54,
+      "grad_norm": 13.433913230895996,
+      "learning_rate": 3.7487487487487488e-06,
+      "loss": 0.4182,
+      "step": 62450
+    },
+    {
+      "epoch": 187.57,
+      "grad_norm": 18.10687828063965,
+      "learning_rate": 3.747747747747748e-06,
+      "loss": 0.3977,
+      "step": 62460
+    },
+    {
+      "epoch": 187.6,
+      "grad_norm": 16.84079933166504,
+      "learning_rate": 3.7467467467467472e-06,
+      "loss": 0.3655,
+      "step": 62470
+    },
+    {
+      "epoch": 187.63,
+      "grad_norm": 21.816804885864258,
+      "learning_rate": 3.745745745745746e-06,
+      "loss": 0.3653,
+      "step": 62480
+    },
+    {
+      "epoch": 187.66,
+      "grad_norm": 17.434526443481445,
+      "learning_rate": 3.744744744744745e-06,
+      "loss": 0.3529,
+      "step": 62490
+    },
+    {
+      "epoch": 187.69,
+      "grad_norm": 11.017670631408691,
+      "learning_rate": 3.743743743743744e-06,
+      "loss": 0.3567,
+      "step": 62500
+    },
+    {
+      "epoch": 187.72,
+      "grad_norm": 15.883646965026855,
+      "learning_rate": 3.7427427427427433e-06,
+      "loss": 0.4037,
+      "step": 62510
+    },
+    {
+      "epoch": 187.75,
+      "grad_norm": 14.018451690673828,
+      "learning_rate": 3.7417417417417423e-06,
+      "loss": 0.3603,
+      "step": 62520
+    },
+    {
+      "epoch": 187.78,
+      "grad_norm": 13.99553394317627,
+      "learning_rate": 3.740740740740741e-06,
+      "loss": 0.3441,
+      "step": 62530
+    },
+    {
+      "epoch": 187.81,
+      "grad_norm": 32.09687423706055,
+      "learning_rate": 3.73973973973974e-06,
+      "loss": 0.3563,
+      "step": 62540
+    },
+    {
+      "epoch": 187.84,
+      "grad_norm": 15.537555694580078,
+      "learning_rate": 3.7387387387387394e-06,
+      "loss": 0.3923,
+      "step": 62550
+    },
+    {
+      "epoch": 187.87,
+      "grad_norm": 14.544533729553223,
+      "learning_rate": 3.7377377377377384e-06,
+      "loss": 0.3755,
+      "step": 62560
+    },
+    {
+      "epoch": 187.9,
+      "grad_norm": 18.86406707763672,
+      "learning_rate": 3.736736736736737e-06,
+      "loss": 0.3809,
+      "step": 62570
+    },
+    {
+      "epoch": 187.93,
+      "grad_norm": 17.87087631225586,
+      "learning_rate": 3.735735735735736e-06,
+      "loss": 0.3728,
+      "step": 62580
+    },
+    {
+      "epoch": 187.96,
+      "grad_norm": 18.669918060302734,
+      "learning_rate": 3.734734734734735e-06,
+      "loss": 0.3856,
+      "step": 62590
+    },
+    {
+      "epoch": 187.99,
+      "grad_norm": 15.052371978759766,
+      "learning_rate": 3.7337337337337345e-06,
+      "loss": 0.416,
+      "step": 62600
+    },
+    {
+      "epoch": 188.0,
+      "eval_accuracy": 0.8621,
+      "eval_loss": 0.535706639289856,
+      "eval_runtime": 12.8389,
+      "eval_samples_per_second": 778.881,
+      "eval_steps_per_second": 3.116,
+      "step": 62604
+    },
+    {
+      "epoch": 188.02,
+      "grad_norm": 17.919443130493164,
+      "learning_rate": 3.732732732732733e-06,
+      "loss": 0.3706,
+      "step": 62610
+    },
+    {
+      "epoch": 188.05,
+      "grad_norm": 15.434700012207031,
+      "learning_rate": 3.731731731731732e-06,
+      "loss": 0.3531,
+      "step": 62620
+    },
+    {
+      "epoch": 188.08,
+      "grad_norm": 17.191574096679688,
+      "learning_rate": 3.730730730730731e-06,
+      "loss": 0.3824,
+      "step": 62630
+    },
+    {
+      "epoch": 188.11,
+      "grad_norm": 11.366107940673828,
+      "learning_rate": 3.72972972972973e-06,
+      "loss": 0.3447,
+      "step": 62640
+    },
+    {
+      "epoch": 188.14,
+      "grad_norm": 14.591341972351074,
+      "learning_rate": 3.7287287287287287e-06,
+      "loss": 0.3185,
+      "step": 62650
+    },
+    {
+      "epoch": 188.17,
+      "grad_norm": 16.664424896240234,
+      "learning_rate": 3.727727727727728e-06,
+      "loss": 0.3383,
+      "step": 62660
+    },
+    {
+      "epoch": 188.2,
+      "grad_norm": 17.518898010253906,
+      "learning_rate": 3.726726726726727e-06,
+      "loss": 0.3288,
+      "step": 62670
+    },
+    {
+      "epoch": 188.23,
+      "grad_norm": 16.329452514648438,
+      "learning_rate": 3.725725725725726e-06,
+      "loss": 0.3553,
+      "step": 62680
+    },
+    {
+      "epoch": 188.26,
+      "grad_norm": 15.844462394714355,
+      "learning_rate": 3.7247247247247248e-06,
+      "loss": 0.375,
+      "step": 62690
+    },
+    {
+      "epoch": 188.29,
+      "grad_norm": 26.54338264465332,
+      "learning_rate": 3.723723723723724e-06,
+      "loss": 0.3644,
+      "step": 62700
+    },
+    {
+      "epoch": 188.32,
+      "grad_norm": 18.89297103881836,
+      "learning_rate": 3.7227227227227232e-06,
+      "loss": 0.3319,
+      "step": 62710
+    },
+    {
+      "epoch": 188.35,
+      "grad_norm": 24.2031192779541,
+      "learning_rate": 3.7217217217217223e-06,
+      "loss": 0.4046,
+      "step": 62720
+    },
+    {
+      "epoch": 188.38,
+      "grad_norm": 24.397083282470703,
+      "learning_rate": 3.720720720720721e-06,
+      "loss": 0.3649,
+      "step": 62730
+    },
+    {
+      "epoch": 188.41,
+      "grad_norm": 13.772913932800293,
+      "learning_rate": 3.71971971971972e-06,
+      "loss": 0.4094,
+      "step": 62740
+    },
+    {
+      "epoch": 188.44,
+      "grad_norm": 21.0062255859375,
+      "learning_rate": 3.718718718718719e-06,
+      "loss": 0.3661,
+      "step": 62750
+    },
+    {
+      "epoch": 188.47,
+      "grad_norm": 12.32347583770752,
+      "learning_rate": 3.7177177177177183e-06,
+      "loss": 0.3615,
+      "step": 62760
+    },
+    {
+      "epoch": 188.5,
+      "grad_norm": 18.984813690185547,
+      "learning_rate": 3.716716716716717e-06,
+      "loss": 0.3559,
+      "step": 62770
+    },
+    {
+      "epoch": 188.53,
+      "grad_norm": 9.388930320739746,
+      "learning_rate": 3.715715715715716e-06,
+      "loss": 0.3401,
+      "step": 62780
+    },
+    {
+      "epoch": 188.56,
+      "grad_norm": 14.62192153930664,
+      "learning_rate": 3.714714714714715e-06,
+      "loss": 0.3882,
+      "step": 62790
+    },
+    {
+      "epoch": 188.59,
+      "grad_norm": 16.153593063354492,
+      "learning_rate": 3.713713713713714e-06,
+      "loss": 0.3305,
+      "step": 62800
+    },
+    {
+      "epoch": 188.62,
+      "grad_norm": 16.790700912475586,
+      "learning_rate": 3.7127127127127134e-06,
+      "loss": 0.3414,
+      "step": 62810
+    },
+    {
+      "epoch": 188.65,
+      "grad_norm": 11.771440505981445,
+      "learning_rate": 3.711711711711712e-06,
+      "loss": 0.3644,
+      "step": 62820
+    },
+    {
+      "epoch": 188.68,
+      "grad_norm": 15.546306610107422,
+      "learning_rate": 3.710710710710711e-06,
+      "loss": 0.4571,
+      "step": 62830
+    },
+    {
+      "epoch": 188.71,
+      "grad_norm": 17.55895233154297,
+      "learning_rate": 3.70970970970971e-06,
+      "loss": 0.3654,
+      "step": 62840
+    },
+    {
+      "epoch": 188.74,
+      "grad_norm": 19.315643310546875,
+      "learning_rate": 3.708708708708709e-06,
+      "loss": 0.4199,
+      "step": 62850
+    },
+    {
+      "epoch": 188.77,
+      "grad_norm": 10.932307243347168,
+      "learning_rate": 3.7077077077077077e-06,
+      "loss": 0.3718,
+      "step": 62860
+    },
+    {
+      "epoch": 188.8,
+      "grad_norm": 11.55249309539795,
+      "learning_rate": 3.706706706706707e-06,
+      "loss": 0.3534,
+      "step": 62870
+    },
+    {
+      "epoch": 188.83,
+      "grad_norm": 24.305089950561523,
+      "learning_rate": 3.705705705705706e-06,
+      "loss": 0.3893,
+      "step": 62880
+    },
+    {
+      "epoch": 188.86,
+      "grad_norm": 12.254450798034668,
+      "learning_rate": 3.704704704704705e-06,
+      "loss": 0.3608,
+      "step": 62890
+    },
+    {
+      "epoch": 188.89,
+      "grad_norm": 16.150676727294922,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.3519,
+      "step": 62900
+    },
+    {
+      "epoch": 188.92,
+      "grad_norm": 23.528825759887695,
+      "learning_rate": 3.7027027027027028e-06,
+      "loss": 0.3941,
+      "step": 62910
+    },
+    {
+      "epoch": 188.95,
+      "grad_norm": 14.541922569274902,
+      "learning_rate": 3.701701701701702e-06,
+      "loss": 0.3322,
+      "step": 62920
+    },
+    {
+      "epoch": 188.98,
+      "grad_norm": 10.297471046447754,
+      "learning_rate": 3.7007007007007012e-06,
+      "loss": 0.3494,
+      "step": 62930
+    },
+    {
+      "epoch": 189.0,
+      "eval_accuracy": 0.864,
+      "eval_loss": 0.5331739783287048,
+      "eval_runtime": 12.7497,
+      "eval_samples_per_second": 784.331,
+      "eval_steps_per_second": 3.137,
+      "step": 62937
+    },
+    {
+      "epoch": 189.01,
+      "grad_norm": 17.502361297607422,
+      "learning_rate": 3.6996996996997e-06,
+      "loss": 0.3487,
+      "step": 62940
+    },
+    {
+      "epoch": 189.04,
+      "grad_norm": 12.237369537353516,
+      "learning_rate": 3.698698698698699e-06,
+      "loss": 0.3586,
+      "step": 62950
+    },
+    {
+      "epoch": 189.07,
+      "grad_norm": 15.576349258422852,
+      "learning_rate": 3.697697697697698e-06,
+      "loss": 0.361,
+      "step": 62960
+    },
+    {
+      "epoch": 189.1,
+      "grad_norm": 13.636961936950684,
+      "learning_rate": 3.6966966966966973e-06,
+      "loss": 0.3612,
+      "step": 62970
+    },
+    {
+      "epoch": 189.13,
+      "grad_norm": 24.699525833129883,
+      "learning_rate": 3.695695695695696e-06,
+      "loss": 0.3409,
+      "step": 62980
+    },
+    {
+      "epoch": 189.16,
+      "grad_norm": 14.128169059753418,
+      "learning_rate": 3.694694694694695e-06,
+      "loss": 0.3657,
+      "step": 62990
+    },
+    {
+      "epoch": 189.19,
+      "grad_norm": 20.979873657226562,
+      "learning_rate": 3.693693693693694e-06,
+      "loss": 0.3528,
+      "step": 63000
+    },
+    {
+      "epoch": 189.22,
+      "grad_norm": 22.696104049682617,
+      "learning_rate": 3.692692692692693e-06,
+      "loss": 0.3434,
+      "step": 63010
+    },
+    {
+      "epoch": 189.25,
+      "grad_norm": 10.990541458129883,
+      "learning_rate": 3.6916916916916915e-06,
+      "loss": 0.3604,
+      "step": 63020
+    },
+    {
+      "epoch": 189.28,
+      "grad_norm": 18.78226089477539,
+      "learning_rate": 3.690690690690691e-06,
+      "loss": 0.3637,
+      "step": 63030
+    },
+    {
+      "epoch": 189.31,
+      "grad_norm": 19.728857040405273,
+      "learning_rate": 3.68968968968969e-06,
+      "loss": 0.3128,
+      "step": 63040
+    },
+    {
+      "epoch": 189.34,
+      "grad_norm": 20.264022827148438,
+      "learning_rate": 3.688688688688689e-06,
+      "loss": 0.3973,
+      "step": 63050
+    },
+    {
+      "epoch": 189.37,
+      "grad_norm": 19.989521026611328,
+      "learning_rate": 3.6876876876876885e-06,
+      "loss": 0.4026,
+      "step": 63060
+    },
+    {
+      "epoch": 189.4,
+      "grad_norm": 14.293108940124512,
+      "learning_rate": 3.6866866866866866e-06,
+      "loss": 0.3195,
+      "step": 63070
+    },
+    {
+      "epoch": 189.43,
+      "grad_norm": 9.645771026611328,
+      "learning_rate": 3.685685685685686e-06,
+      "loss": 0.3727,
+      "step": 63080
+    },
+    {
+      "epoch": 189.46,
+      "grad_norm": 11.41912841796875,
+      "learning_rate": 3.684684684684685e-06,
+      "loss": 0.3659,
+      "step": 63090
+    },
+    {
+      "epoch": 189.49,
+      "grad_norm": 23.68900489807129,
+      "learning_rate": 3.683683683683684e-06,
+      "loss": 0.3222,
+      "step": 63100
+    },
+    {
+      "epoch": 189.52,
+      "grad_norm": 14.355560302734375,
+      "learning_rate": 3.6826826826826827e-06,
+      "loss": 0.3147,
+      "step": 63110
+    },
+    {
+      "epoch": 189.55,
+      "grad_norm": 29.992033004760742,
+      "learning_rate": 3.681681681681682e-06,
+      "loss": 0.3691,
+      "step": 63120
+    },
+    {
+      "epoch": 189.58,
+      "grad_norm": 15.50009536743164,
+      "learning_rate": 3.680680680680681e-06,
+      "loss": 0.3794,
+      "step": 63130
+    },
+    {
+      "epoch": 189.61,
+      "grad_norm": 14.663280487060547,
+      "learning_rate": 3.67967967967968e-06,
+      "loss": 0.3583,
+      "step": 63140
+    },
+    {
+      "epoch": 189.64,
+      "grad_norm": 13.728165626525879,
+      "learning_rate": 3.6786786786786788e-06,
+      "loss": 0.3679,
+      "step": 63150
+    },
+    {
+      "epoch": 189.67,
+      "grad_norm": 12.170446395874023,
+      "learning_rate": 3.677677677677678e-06,
+      "loss": 0.3558,
+      "step": 63160
+    },
+    {
+      "epoch": 189.7,
+      "grad_norm": 24.05000114440918,
+      "learning_rate": 3.6766766766766772e-06,
+      "loss": 0.3012,
+      "step": 63170
+    },
+    {
+      "epoch": 189.73,
+      "grad_norm": 11.954524040222168,
+      "learning_rate": 3.6756756756756763e-06,
+      "loss": 0.3704,
+      "step": 63180
+    },
+    {
+      "epoch": 189.76,
+      "grad_norm": 7.776760101318359,
+      "learning_rate": 3.674674674674675e-06,
+      "loss": 0.3281,
+      "step": 63190
+    },
+    {
+      "epoch": 189.79,
+      "grad_norm": 18.735029220581055,
+      "learning_rate": 3.673673673673674e-06,
+      "loss": 0.3419,
+      "step": 63200
+    },
+    {
+      "epoch": 189.82,
+      "grad_norm": 25.53712272644043,
+      "learning_rate": 3.672672672672673e-06,
+      "loss": 0.3849,
+      "step": 63210
+    },
+    {
+      "epoch": 189.85,
+      "grad_norm": 21.539722442626953,
+      "learning_rate": 3.6716716716716723e-06,
+      "loss": 0.3656,
+      "step": 63220
+    },
+    {
+      "epoch": 189.88,
+      "grad_norm": 15.000730514526367,
+      "learning_rate": 3.670670670670671e-06,
+      "loss": 0.4458,
+      "step": 63230
+    },
+    {
+      "epoch": 189.91,
+      "grad_norm": 17.676252365112305,
+      "learning_rate": 3.66966966966967e-06,
+      "loss": 0.34,
+      "step": 63240
+    },
+    {
+      "epoch": 189.94,
+      "grad_norm": 26.982040405273438,
+      "learning_rate": 3.668668668668669e-06,
+      "loss": 0.3856,
+      "step": 63250
+    },
+    {
+      "epoch": 189.97,
+      "grad_norm": 16.351024627685547,
+      "learning_rate": 3.667667667667668e-06,
+      "loss": 0.3366,
+      "step": 63260
+    },
+    {
+      "epoch": 190.0,
+      "grad_norm": 160.1409454345703,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 0.4009,
+      "step": 63270
+    },
+    {
+      "epoch": 190.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.5363979339599609,
+      "eval_runtime": 12.721,
+      "eval_samples_per_second": 786.103,
+      "eval_steps_per_second": 3.144,
+      "step": 63270
+    },
+    {
+      "epoch": 190.03,
+      "grad_norm": 18.11351776123047,
+      "learning_rate": 3.665665665665666e-06,
+      "loss": 0.3443,
+      "step": 63280
+    },
+    {
+      "epoch": 190.06,
+      "grad_norm": 11.389575958251953,
+      "learning_rate": 3.664664664664665e-06,
+      "loss": 0.3509,
+      "step": 63290
+    },
+    {
+      "epoch": 190.09,
+      "grad_norm": 13.321301460266113,
+      "learning_rate": 3.663663663663664e-06,
+      "loss": 0.3719,
+      "step": 63300
+    },
+    {
+      "epoch": 190.12,
+      "grad_norm": 10.897168159484863,
+      "learning_rate": 3.662662662662663e-06,
+      "loss": 0.3607,
+      "step": 63310
+    },
+    {
+      "epoch": 190.15,
+      "grad_norm": 13.854853630065918,
+      "learning_rate": 3.6616616616616617e-06,
+      "loss": 0.3614,
+      "step": 63320
+    },
+    {
+      "epoch": 190.18,
+      "grad_norm": 16.602008819580078,
+      "learning_rate": 3.660660660660661e-06,
+      "loss": 0.3796,
+      "step": 63330
+    },
+    {
+      "epoch": 190.21,
+      "grad_norm": 12.238582611083984,
+      "learning_rate": 3.65965965965966e-06,
+      "loss": 0.371,
+      "step": 63340
+    },
+    {
+      "epoch": 190.24,
+      "grad_norm": 17.719486236572266,
+      "learning_rate": 3.658658658658659e-06,
+      "loss": 0.3955,
+      "step": 63350
+    },
+    {
+      "epoch": 190.27,
+      "grad_norm": 14.483688354492188,
+      "learning_rate": 3.6576576576576577e-06,
+      "loss": 0.4154,
+      "step": 63360
+    },
+    {
+      "epoch": 190.3,
+      "grad_norm": 13.531072616577148,
+      "learning_rate": 3.6566566566566568e-06,
+      "loss": 0.3778,
+      "step": 63370
+    },
+    {
+      "epoch": 190.33,
+      "grad_norm": 10.521717071533203,
+      "learning_rate": 3.655655655655656e-06,
+      "loss": 0.3447,
+      "step": 63380
+    },
+    {
+      "epoch": 190.36,
+      "grad_norm": 16.310861587524414,
+      "learning_rate": 3.6546546546546552e-06,
+      "loss": 0.329,
+      "step": 63390
+    },
+    {
+      "epoch": 190.39,
+      "grad_norm": 20.474990844726562,
+      "learning_rate": 3.653653653653654e-06,
+      "loss": 0.4047,
+      "step": 63400
+    },
+    {
+      "epoch": 190.42,
+      "grad_norm": 14.680219650268555,
+      "learning_rate": 3.652652652652653e-06,
+      "loss": 0.3735,
+      "step": 63410
+    },
+    {
+      "epoch": 190.45,
+      "grad_norm": 19.08514976501465,
+      "learning_rate": 3.651651651651652e-06,
+      "loss": 0.3361,
+      "step": 63420
+    },
+    {
+      "epoch": 190.48,
+      "grad_norm": 14.072154998779297,
+      "learning_rate": 3.6506506506506513e-06,
+      "loss": 0.3566,
+      "step": 63430
+    },
+    {
+      "epoch": 190.51,
+      "grad_norm": 13.485198974609375,
+      "learning_rate": 3.64964964964965e-06,
+      "loss": 0.3711,
+      "step": 63440
+    },
+    {
+      "epoch": 190.54,
+      "grad_norm": 15.185403823852539,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.3142,
+      "step": 63450
+    },
+    {
+      "epoch": 190.57,
+      "grad_norm": 22.789209365844727,
+      "learning_rate": 3.647647647647648e-06,
+      "loss": 0.3459,
+      "step": 63460
+    },
+    {
+      "epoch": 190.6,
+      "grad_norm": 20.936811447143555,
+      "learning_rate": 3.646646646646647e-06,
+      "loss": 0.371,
+      "step": 63470
+    },
+    {
+      "epoch": 190.63,
+      "grad_norm": 17.250802993774414,
+      "learning_rate": 3.6456456456456455e-06,
+      "loss": 0.441,
+      "step": 63480
+    },
+    {
+      "epoch": 190.66,
+      "grad_norm": 13.7217435836792,
+      "learning_rate": 3.644644644644645e-06,
+      "loss": 0.3279,
+      "step": 63490
+    },
+    {
+      "epoch": 190.69,
+      "grad_norm": 12.822487831115723,
+      "learning_rate": 3.643643643643644e-06,
+      "loss": 0.3401,
+      "step": 63500
+    },
+    {
+      "epoch": 190.72,
+      "grad_norm": 14.110260963439941,
+      "learning_rate": 3.642642642642643e-06,
+      "loss": 0.343,
+      "step": 63510
+    },
+    {
+      "epoch": 190.75,
+      "grad_norm": 22.540386199951172,
+      "learning_rate": 3.6416416416416416e-06,
+      "loss": 0.351,
+      "step": 63520
+    },
+    {
+      "epoch": 190.78,
+      "grad_norm": 17.715463638305664,
+      "learning_rate": 3.6406406406406406e-06,
+      "loss": 0.3699,
+      "step": 63530
+    },
+    {
+      "epoch": 190.81,
+      "grad_norm": 17.999818801879883,
+      "learning_rate": 3.63963963963964e-06,
+      "loss": 0.3366,
+      "step": 63540
+    },
+    {
+      "epoch": 190.84,
+      "grad_norm": 13.149312973022461,
+      "learning_rate": 3.638638638638639e-06,
+      "loss": 0.4037,
+      "step": 63550
+    },
+    {
+      "epoch": 190.87,
+      "grad_norm": 26.187606811523438,
+      "learning_rate": 3.6376376376376377e-06,
+      "loss": 0.3923,
+      "step": 63560
+    },
+    {
+      "epoch": 190.9,
+      "grad_norm": 16.476242065429688,
+      "learning_rate": 3.6366366366366367e-06,
+      "loss": 0.3582,
+      "step": 63570
+    },
+    {
+      "epoch": 190.93,
+      "grad_norm": 14.544086456298828,
+      "learning_rate": 3.635635635635636e-06,
+      "loss": 0.3289,
+      "step": 63580
+    },
+    {
+      "epoch": 190.96,
+      "grad_norm": 15.085927963256836,
+      "learning_rate": 3.634634634634635e-06,
+      "loss": 0.3969,
+      "step": 63590
+    },
+    {
+      "epoch": 190.99,
+      "grad_norm": 11.291979789733887,
+      "learning_rate": 3.633633633633634e-06,
+      "loss": 0.3935,
+      "step": 63600
+    },
+    {
+      "epoch": 191.0,
+      "eval_accuracy": 0.8668,
+      "eval_loss": 0.5328532457351685,
+      "eval_runtime": 12.5422,
+      "eval_samples_per_second": 797.309,
+      "eval_steps_per_second": 3.189,
+      "step": 63603
+    },
+    {
+      "epoch": 191.02,
+      "grad_norm": 16.855735778808594,
+      "learning_rate": 3.6326326326326328e-06,
+      "loss": 0.4076,
+      "step": 63610
+    },
+    {
+      "epoch": 191.05,
+      "grad_norm": 18.80243682861328,
+      "learning_rate": 3.631631631631632e-06,
+      "loss": 0.3396,
+      "step": 63620
+    },
+    {
+      "epoch": 191.08,
+      "grad_norm": 16.511320114135742,
+      "learning_rate": 3.6306306306306312e-06,
+      "loss": 0.365,
+      "step": 63630
+    },
+    {
+      "epoch": 191.11,
+      "grad_norm": 16.43462562561035,
+      "learning_rate": 3.6296296296296302e-06,
+      "loss": 0.3327,
+      "step": 63640
+    },
+    {
+      "epoch": 191.14,
+      "grad_norm": 16.016698837280273,
+      "learning_rate": 3.628628628628629e-06,
+      "loss": 0.3631,
+      "step": 63650
+    },
+    {
+      "epoch": 191.17,
+      "grad_norm": 13.91424560546875,
+      "learning_rate": 3.627627627627628e-06,
+      "loss": 0.2966,
+      "step": 63660
+    },
+    {
+      "epoch": 191.2,
+      "grad_norm": 11.636504173278809,
+      "learning_rate": 3.626626626626627e-06,
+      "loss": 0.3951,
+      "step": 63670
+    },
+    {
+      "epoch": 191.23,
+      "grad_norm": 15.57904052734375,
+      "learning_rate": 3.6256256256256263e-06,
+      "loss": 0.429,
+      "step": 63680
+    },
+    {
+      "epoch": 191.26,
+      "grad_norm": 12.698409080505371,
+      "learning_rate": 3.624624624624625e-06,
+      "loss": 0.372,
+      "step": 63690
+    },
+    {
+      "epoch": 191.29,
+      "grad_norm": 17.87053108215332,
+      "learning_rate": 3.623623623623624e-06,
+      "loss": 0.3413,
+      "step": 63700
+    },
+    {
+      "epoch": 191.32,
+      "grad_norm": 22.22568702697754,
+      "learning_rate": 3.622622622622623e-06,
+      "loss": 0.3624,
+      "step": 63710
+    },
+    {
+      "epoch": 191.35,
+      "grad_norm": 16.12151336669922,
+      "learning_rate": 3.621621621621622e-06,
+      "loss": 0.3642,
+      "step": 63720
+    },
+    {
+      "epoch": 191.38,
+      "grad_norm": 13.774282455444336,
+      "learning_rate": 3.6206206206206206e-06,
+      "loss": 0.398,
+      "step": 63730
+    },
+    {
+      "epoch": 191.41,
+      "grad_norm": 15.622909545898438,
+      "learning_rate": 3.61961961961962e-06,
+      "loss": 0.3672,
+      "step": 63740
+    },
+    {
+      "epoch": 191.44,
+      "grad_norm": 18.793437957763672,
+      "learning_rate": 3.618618618618619e-06,
+      "loss": 0.374,
+      "step": 63750
+    },
+    {
+      "epoch": 191.47,
+      "grad_norm": 15.699793815612793,
+      "learning_rate": 3.617617617617618e-06,
+      "loss": 0.3326,
+      "step": 63760
+    },
+    {
+      "epoch": 191.5,
+      "grad_norm": 26.408864974975586,
+      "learning_rate": 3.6166166166166166e-06,
+      "loss": 0.3982,
+      "step": 63770
+    },
+    {
+      "epoch": 191.53,
+      "grad_norm": 12.821403503417969,
+      "learning_rate": 3.6156156156156157e-06,
+      "loss": 0.3704,
+      "step": 63780
+    },
+    {
+      "epoch": 191.56,
+      "grad_norm": 14.383544921875,
+      "learning_rate": 3.614614614614615e-06,
+      "loss": 0.3448,
+      "step": 63790
+    },
+    {
+      "epoch": 191.59,
+      "grad_norm": 13.512648582458496,
+      "learning_rate": 3.613613613613614e-06,
+      "loss": 0.3989,
+      "step": 63800
+    },
+    {
+      "epoch": 191.62,
+      "grad_norm": 19.62908172607422,
+      "learning_rate": 3.6126126126126127e-06,
+      "loss": 0.3678,
+      "step": 63810
+    },
+    {
+      "epoch": 191.65,
+      "grad_norm": 24.276885986328125,
+      "learning_rate": 3.6116116116116117e-06,
+      "loss": 0.3495,
+      "step": 63820
+    },
+    {
+      "epoch": 191.68,
+      "grad_norm": 14.500049591064453,
+      "learning_rate": 3.6106106106106108e-06,
+      "loss": 0.3261,
+      "step": 63830
+    },
+    {
+      "epoch": 191.71,
+      "grad_norm": 14.278578758239746,
+      "learning_rate": 3.60960960960961e-06,
+      "loss": 0.3518,
+      "step": 63840
+    },
+    {
+      "epoch": 191.74,
+      "grad_norm": 17.048242568969727,
+      "learning_rate": 3.608608608608609e-06,
+      "loss": 0.3747,
+      "step": 63850
+    },
+    {
+      "epoch": 191.77,
+      "grad_norm": 15.536795616149902,
+      "learning_rate": 3.607607607607608e-06,
+      "loss": 0.3484,
+      "step": 63860
+    },
+    {
+      "epoch": 191.8,
+      "grad_norm": 11.230851173400879,
+      "learning_rate": 3.606606606606607e-06,
+      "loss": 0.3129,
+      "step": 63870
+    },
+    {
+      "epoch": 191.83,
+      "grad_norm": 12.430225372314453,
+      "learning_rate": 3.605605605605606e-06,
+      "loss": 0.3848,
+      "step": 63880
+    },
+    {
+      "epoch": 191.86,
+      "grad_norm": 13.873812675476074,
+      "learning_rate": 3.6046046046046053e-06,
+      "loss": 0.3537,
+      "step": 63890
+    },
+    {
+      "epoch": 191.89,
+      "grad_norm": 17.284290313720703,
+      "learning_rate": 3.603603603603604e-06,
+      "loss": 0.3539,
+      "step": 63900
+    },
+    {
+      "epoch": 191.92,
+      "grad_norm": 10.284464836120605,
+      "learning_rate": 3.602602602602603e-06,
+      "loss": 0.3376,
+      "step": 63910
+    },
+    {
+      "epoch": 191.95,
+      "grad_norm": 15.662520408630371,
+      "learning_rate": 3.601601601601602e-06,
+      "loss": 0.3856,
+      "step": 63920
+    },
+    {
+      "epoch": 191.98,
+      "grad_norm": 16.96831703186035,
+      "learning_rate": 3.600600600600601e-06,
+      "loss": 0.3666,
+      "step": 63930
+    },
+    {
+      "epoch": 192.0,
+      "eval_accuracy": 0.8641,
+      "eval_loss": 0.5337411761283875,
+      "eval_runtime": 12.9596,
+      "eval_samples_per_second": 771.628,
+      "eval_steps_per_second": 3.087,
+      "step": 63936
+    },
+    {
+      "epoch": 192.01,
+      "grad_norm": 15.058183670043945,
+      "learning_rate": 3.5995995995995995e-06,
+      "loss": 0.368,
+      "step": 63940
+    },
+    {
+      "epoch": 192.04,
+      "grad_norm": 11.408941268920898,
+      "learning_rate": 3.598598598598599e-06,
+      "loss": 0.4045,
+      "step": 63950
+    },
+    {
+      "epoch": 192.07,
+      "grad_norm": 14.189217567443848,
+      "learning_rate": 3.597597597597598e-06,
+      "loss": 0.3898,
+      "step": 63960
+    },
+    {
+      "epoch": 192.1,
+      "grad_norm": 14.804679870605469,
+      "learning_rate": 3.596596596596597e-06,
+      "loss": 0.3718,
+      "step": 63970
+    },
+    {
+      "epoch": 192.13,
+      "grad_norm": 13.726759910583496,
+      "learning_rate": 3.5955955955955956e-06,
+      "loss": 0.3511,
+      "step": 63980
+    },
+    {
+      "epoch": 192.16,
+      "grad_norm": 16.841888427734375,
+      "learning_rate": 3.5945945945945946e-06,
+      "loss": 0.3777,
+      "step": 63990
+    },
+    {
+      "epoch": 192.19,
+      "grad_norm": 12.246529579162598,
+      "learning_rate": 3.593593593593594e-06,
+      "loss": 0.3055,
+      "step": 64000
+    },
+    {
+      "epoch": 192.22,
+      "grad_norm": 15.01268482208252,
+      "learning_rate": 3.592592592592593e-06,
+      "loss": 0.3249,
+      "step": 64010
+    },
+    {
+      "epoch": 192.25,
+      "grad_norm": 16.718412399291992,
+      "learning_rate": 3.5915915915915917e-06,
+      "loss": 0.3269,
+      "step": 64020
+    },
+    {
+      "epoch": 192.28,
+      "grad_norm": 12.279858589172363,
+      "learning_rate": 3.5905905905905907e-06,
+      "loss": 0.3267,
+      "step": 64030
+    },
+    {
+      "epoch": 192.31,
+      "grad_norm": 16.63195037841797,
+      "learning_rate": 3.5895895895895897e-06,
+      "loss": 0.3303,
+      "step": 64040
+    },
+    {
+      "epoch": 192.34,
+      "grad_norm": 21.630779266357422,
+      "learning_rate": 3.588588588588589e-06,
+      "loss": 0.4151,
+      "step": 64050
+    },
+    {
+      "epoch": 192.37,
+      "grad_norm": 10.962024688720703,
+      "learning_rate": 3.5875875875875877e-06,
+      "loss": 0.3667,
+      "step": 64060
+    },
+    {
+      "epoch": 192.4,
+      "grad_norm": 12.721113204956055,
+      "learning_rate": 3.5865865865865868e-06,
+      "loss": 0.3608,
+      "step": 64070
+    },
+    {
+      "epoch": 192.43,
+      "grad_norm": 10.111197471618652,
+      "learning_rate": 3.5855855855855858e-06,
+      "loss": 0.3671,
+      "step": 64080
+    },
+    {
+      "epoch": 192.46,
+      "grad_norm": 17.345800399780273,
+      "learning_rate": 3.5845845845845852e-06,
+      "loss": 0.3677,
+      "step": 64090
+    },
+    {
+      "epoch": 192.49,
+      "grad_norm": 22.25749397277832,
+      "learning_rate": 3.5835835835835834e-06,
+      "loss": 0.3679,
+      "step": 64100
+    },
+    {
+      "epoch": 192.52,
+      "grad_norm": 12.035635948181152,
+      "learning_rate": 3.582582582582583e-06,
+      "loss": 0.3796,
+      "step": 64110
+    },
+    {
+      "epoch": 192.55,
+      "grad_norm": 13.641653060913086,
+      "learning_rate": 3.581581581581582e-06,
+      "loss": 0.3177,
+      "step": 64120
+    },
+    {
+      "epoch": 192.58,
+      "grad_norm": 17.43212890625,
+      "learning_rate": 3.580580580580581e-06,
+      "loss": 0.379,
+      "step": 64130
+    },
+    {
+      "epoch": 192.61,
+      "grad_norm": 14.521151542663574,
+      "learning_rate": 3.5795795795795803e-06,
+      "loss": 0.3194,
+      "step": 64140
+    },
+    {
+      "epoch": 192.64,
+      "grad_norm": 15.088191986083984,
+      "learning_rate": 3.578578578578579e-06,
+      "loss": 0.3652,
+      "step": 64150
+    },
+    {
+      "epoch": 192.67,
+      "grad_norm": 18.062393188476562,
+      "learning_rate": 3.577577577577578e-06,
+      "loss": 0.3796,
+      "step": 64160
+    },
+    {
+      "epoch": 192.7,
+      "grad_norm": 19.616348266601562,
+      "learning_rate": 3.576576576576577e-06,
+      "loss": 0.3298,
+      "step": 64170
+    },
+    {
+      "epoch": 192.73,
+      "grad_norm": 9.022031784057617,
+      "learning_rate": 3.575575575575576e-06,
+      "loss": 0.3643,
+      "step": 64180
+    },
+    {
+      "epoch": 192.76,
+      "grad_norm": 19.191675186157227,
+      "learning_rate": 3.5745745745745746e-06,
+      "loss": 0.3495,
+      "step": 64190
+    },
+    {
+      "epoch": 192.79,
+      "grad_norm": 14.186291694641113,
+      "learning_rate": 3.573573573573574e-06,
+      "loss": 0.3948,
+      "step": 64200
+    },
+    {
+      "epoch": 192.82,
+      "grad_norm": 16.721712112426758,
+      "learning_rate": 3.572572572572573e-06,
+      "loss": 0.3716,
+      "step": 64210
+    },
+    {
+      "epoch": 192.85,
+      "grad_norm": 19.428600311279297,
+      "learning_rate": 3.571571571571572e-06,
+      "loss": 0.3774,
+      "step": 64220
+    },
+    {
+      "epoch": 192.88,
+      "grad_norm": 18.23359489440918,
+      "learning_rate": 3.5705705705705706e-06,
+      "loss": 0.3797,
+      "step": 64230
+    },
+    {
+      "epoch": 192.91,
+      "grad_norm": 17.599369049072266,
+      "learning_rate": 3.5695695695695697e-06,
+      "loss": 0.3597,
+      "step": 64240
+    },
+    {
+      "epoch": 192.94,
+      "grad_norm": 14.374923706054688,
+      "learning_rate": 3.568568568568569e-06,
+      "loss": 0.4093,
+      "step": 64250
+    },
+    {
+      "epoch": 192.97,
+      "grad_norm": 13.950732231140137,
+      "learning_rate": 3.567567567567568e-06,
+      "loss": 0.3474,
+      "step": 64260
+    },
+    {
+      "epoch": 193.0,
+      "eval_accuracy": 0.866,
+      "eval_loss": 0.5321243405342102,
+      "eval_runtime": 12.9186,
+      "eval_samples_per_second": 774.079,
+      "eval_steps_per_second": 3.096,
+      "step": 64269
+    },
+    {
+      "epoch": 193.0,
+      "grad_norm": 16.014907836914062,
+      "learning_rate": 3.5665665665665667e-06,
+      "loss": 0.3657,
+      "step": 64270
+    },
+    {
+      "epoch": 193.03,
+      "grad_norm": 19.22174644470215,
+      "learning_rate": 3.5655655655655657e-06,
+      "loss": 0.4043,
+      "step": 64280
+    },
+    {
+      "epoch": 193.06,
+      "grad_norm": 16.375764846801758,
+      "learning_rate": 3.5645645645645647e-06,
+      "loss": 0.3561,
+      "step": 64290
+    },
+    {
+      "epoch": 193.09,
+      "grad_norm": 11.03246784210205,
+      "learning_rate": 3.563563563563564e-06,
+      "loss": 0.3907,
+      "step": 64300
+    },
+    {
+      "epoch": 193.12,
+      "grad_norm": 10.557662010192871,
+      "learning_rate": 3.5625625625625628e-06,
+      "loss": 0.4154,
+      "step": 64310
+    },
+    {
+      "epoch": 193.15,
+      "grad_norm": 17.667743682861328,
+      "learning_rate": 3.561561561561562e-06,
+      "loss": 0.3285,
+      "step": 64320
+    },
+    {
+      "epoch": 193.18,
+      "grad_norm": 16.66663932800293,
+      "learning_rate": 3.560560560560561e-06,
+      "loss": 0.3164,
+      "step": 64330
+    },
+    {
+      "epoch": 193.21,
+      "grad_norm": 16.536239624023438,
+      "learning_rate": 3.55955955955956e-06,
+      "loss": 0.3474,
+      "step": 64340
+    },
+    {
+      "epoch": 193.24,
+      "grad_norm": 17.186443328857422,
+      "learning_rate": 3.5585585585585584e-06,
+      "loss": 0.4144,
+      "step": 64350
+    },
+    {
+      "epoch": 193.27,
+      "grad_norm": 10.556830406188965,
+      "learning_rate": 3.557557557557558e-06,
+      "loss": 0.3335,
+      "step": 64360
+    },
+    {
+      "epoch": 193.3,
+      "grad_norm": 15.08750057220459,
+      "learning_rate": 3.556556556556557e-06,
+      "loss": 0.3535,
+      "step": 64370
+    },
+    {
+      "epoch": 193.33,
+      "grad_norm": 16.936471939086914,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 0.348,
+      "step": 64380
+    },
+    {
+      "epoch": 193.36,
+      "grad_norm": 17.044754028320312,
+      "learning_rate": 3.554554554554555e-06,
+      "loss": 0.3439,
+      "step": 64390
+    },
+    {
+      "epoch": 193.39,
+      "grad_norm": 12.411555290222168,
+      "learning_rate": 3.5535535535535535e-06,
+      "loss": 0.3307,
+      "step": 64400
+    },
+    {
+      "epoch": 193.42,
+      "grad_norm": 20.38922119140625,
+      "learning_rate": 3.552552552552553e-06,
+      "loss": 0.359,
+      "step": 64410
+    },
+    {
+      "epoch": 193.45,
+      "grad_norm": 11.485054016113281,
+      "learning_rate": 3.551551551551552e-06,
+      "loss": 0.3688,
+      "step": 64420
+    },
+    {
+      "epoch": 193.48,
+      "grad_norm": 12.716190338134766,
+      "learning_rate": 3.550550550550551e-06,
+      "loss": 0.3707,
+      "step": 64430
+    },
+    {
+      "epoch": 193.51,
+      "grad_norm": 17.378948211669922,
+      "learning_rate": 3.5495495495495496e-06,
+      "loss": 0.3607,
+      "step": 64440
+    },
+    {
+      "epoch": 193.54,
+      "grad_norm": 16.649152755737305,
+      "learning_rate": 3.5485485485485486e-06,
+      "loss": 0.3319,
+      "step": 64450
+    },
+    {
+      "epoch": 193.57,
+      "grad_norm": 13.961941719055176,
+      "learning_rate": 3.547547547547548e-06,
+      "loss": 0.3499,
+      "step": 64460
+    },
+    {
+      "epoch": 193.6,
+      "grad_norm": 10.565949440002441,
+      "learning_rate": 3.546546546546547e-06,
+      "loss": 0.3273,
+      "step": 64470
+    },
+    {
+      "epoch": 193.63,
+      "grad_norm": 12.998795509338379,
+      "learning_rate": 3.5455455455455457e-06,
+      "loss": 0.3976,
+      "step": 64480
+    },
+    {
+      "epoch": 193.66,
+      "grad_norm": 17.00520133972168,
+      "learning_rate": 3.5445445445445447e-06,
+      "loss": 0.3698,
+      "step": 64490
+    },
+    {
+      "epoch": 193.69,
+      "grad_norm": 14.313894271850586,
+      "learning_rate": 3.5435435435435437e-06,
+      "loss": 0.4027,
+      "step": 64500
+    },
+    {
+      "epoch": 193.72,
+      "grad_norm": 16.678760528564453,
+      "learning_rate": 3.542542542542543e-06,
+      "loss": 0.3334,
+      "step": 64510
+    },
+    {
+      "epoch": 193.75,
+      "grad_norm": 16.32000732421875,
+      "learning_rate": 3.5415415415415417e-06,
+      "loss": 0.3477,
+      "step": 64520
+    },
+    {
+      "epoch": 193.78,
+      "grad_norm": 19.507221221923828,
+      "learning_rate": 3.5405405405405408e-06,
+      "loss": 0.3425,
+      "step": 64530
+    },
+    {
+      "epoch": 193.81,
+      "grad_norm": 20.301767349243164,
+      "learning_rate": 3.5395395395395398e-06,
+      "loss": 0.35,
+      "step": 64540
+    },
+    {
+      "epoch": 193.84,
+      "grad_norm": 17.980737686157227,
+      "learning_rate": 3.5385385385385392e-06,
+      "loss": 0.3402,
+      "step": 64550
+    },
+    {
+      "epoch": 193.87,
+      "grad_norm": 15.32487964630127,
+      "learning_rate": 3.5375375375375374e-06,
+      "loss": 0.3588,
+      "step": 64560
+    },
+    {
+      "epoch": 193.9,
+      "grad_norm": 17.909578323364258,
+      "learning_rate": 3.536536536536537e-06,
+      "loss": 0.3777,
+      "step": 64570
+    },
+    {
+      "epoch": 193.93,
+      "grad_norm": 18.32218360900879,
+      "learning_rate": 3.535535535535536e-06,
+      "loss": 0.3894,
+      "step": 64580
+    },
+    {
+      "epoch": 193.96,
+      "grad_norm": 17.508325576782227,
+      "learning_rate": 3.534534534534535e-06,
+      "loss": 0.3642,
+      "step": 64590
+    },
+    {
+      "epoch": 193.99,
+      "grad_norm": 13.325234413146973,
+      "learning_rate": 3.5335335335335335e-06,
+      "loss": 0.3873,
+      "step": 64600
+    },
+    {
+      "epoch": 194.0,
+      "eval_accuracy": 0.8635,
+      "eval_loss": 0.5335552096366882,
+      "eval_runtime": 12.7421,
+      "eval_samples_per_second": 784.798,
+      "eval_steps_per_second": 3.139,
+      "step": 64602
+    },
+    {
+      "epoch": 194.02,
+      "grad_norm": 25.10406494140625,
+      "learning_rate": 3.532532532532533e-06,
+      "loss": 0.3931,
+      "step": 64610
+    },
+    {
+      "epoch": 194.05,
+      "grad_norm": 12.969768524169922,
+      "learning_rate": 3.531531531531532e-06,
+      "loss": 0.3148,
+      "step": 64620
+    },
+    {
+      "epoch": 194.08,
+      "grad_norm": 15.763026237487793,
+      "learning_rate": 3.530530530530531e-06,
+      "loss": 0.3892,
+      "step": 64630
+    },
+    {
+      "epoch": 194.11,
+      "grad_norm": 13.997791290283203,
+      "learning_rate": 3.5295295295295295e-06,
+      "loss": 0.3642,
+      "step": 64640
+    },
+    {
+      "epoch": 194.14,
+      "grad_norm": 17.048036575317383,
+      "learning_rate": 3.5285285285285286e-06,
+      "loss": 0.3718,
+      "step": 64650
+    },
+    {
+      "epoch": 194.17,
+      "grad_norm": 11.98378849029541,
+      "learning_rate": 3.527527527527528e-06,
+      "loss": 0.3973,
+      "step": 64660
+    },
+    {
+      "epoch": 194.2,
+      "grad_norm": 19.660776138305664,
+      "learning_rate": 3.526526526526527e-06,
+      "loss": 0.3399,
+      "step": 64670
+    },
+    {
+      "epoch": 194.23,
+      "grad_norm": 8.274032592773438,
+      "learning_rate": 3.525525525525526e-06,
+      "loss": 0.3398,
+      "step": 64680
+    },
+    {
+      "epoch": 194.26,
+      "grad_norm": 17.23844337463379,
+      "learning_rate": 3.5245245245245246e-06,
+      "loss": 0.3117,
+      "step": 64690
+    },
+    {
+      "epoch": 194.29,
+      "grad_norm": 18.12611198425293,
+      "learning_rate": 3.5235235235235237e-06,
+      "loss": 0.3701,
+      "step": 64700
+    },
+    {
+      "epoch": 194.32,
+      "grad_norm": 13.265703201293945,
+      "learning_rate": 3.522522522522523e-06,
+      "loss": 0.3681,
+      "step": 64710
+    },
+    {
+      "epoch": 194.35,
+      "grad_norm": 15.62252426147461,
+      "learning_rate": 3.521521521521522e-06,
+      "loss": 0.3369,
+      "step": 64720
+    },
+    {
+      "epoch": 194.38,
+      "grad_norm": 20.061342239379883,
+      "learning_rate": 3.5205205205205207e-06,
+      "loss": 0.3403,
+      "step": 64730
+    },
+    {
+      "epoch": 194.41,
+      "grad_norm": 18.891294479370117,
+      "learning_rate": 3.5195195195195197e-06,
+      "loss": 0.3377,
+      "step": 64740
+    },
+    {
+      "epoch": 194.44,
+      "grad_norm": 13.411861419677734,
+      "learning_rate": 3.5185185185185187e-06,
+      "loss": 0.4382,
+      "step": 64750
+    },
+    {
+      "epoch": 194.47,
+      "grad_norm": 16.1318416595459,
+      "learning_rate": 3.517517517517518e-06,
+      "loss": 0.3503,
+      "step": 64760
+    },
+    {
+      "epoch": 194.5,
+      "grad_norm": 12.190544128417969,
+      "learning_rate": 3.5165165165165168e-06,
+      "loss": 0.335,
+      "step": 64770
+    },
+    {
+      "epoch": 194.53,
+      "grad_norm": 13.74680233001709,
+      "learning_rate": 3.515515515515516e-06,
+      "loss": 0.3392,
+      "step": 64780
+    },
+    {
+      "epoch": 194.56,
+      "grad_norm": 31.573347091674805,
+      "learning_rate": 3.514514514514515e-06,
+      "loss": 0.35,
+      "step": 64790
+    },
+    {
+      "epoch": 194.59,
+      "grad_norm": 16.8371524810791,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.3975,
+      "step": 64800
+    },
+    {
+      "epoch": 194.62,
+      "grad_norm": 14.153903007507324,
+      "learning_rate": 3.5125125125125124e-06,
+      "loss": 0.3562,
+      "step": 64810
+    },
+    {
+      "epoch": 194.65,
+      "grad_norm": 22.134521484375,
+      "learning_rate": 3.511511511511512e-06,
+      "loss": 0.3542,
+      "step": 64820
+    },
+    {
+      "epoch": 194.68,
+      "grad_norm": 13.255815505981445,
+      "learning_rate": 3.510510510510511e-06,
+      "loss": 0.3296,
+      "step": 64830
+    },
+    {
+      "epoch": 194.71,
+      "grad_norm": 11.560598373413086,
+      "learning_rate": 3.50950950950951e-06,
+      "loss": 0.3504,
+      "step": 64840
+    },
+    {
+      "epoch": 194.74,
+      "grad_norm": 11.758158683776855,
+      "learning_rate": 3.5085085085085085e-06,
+      "loss": 0.3882,
+      "step": 64850
+    },
+    {
+      "epoch": 194.77,
+      "grad_norm": 11.432168960571289,
+      "learning_rate": 3.5075075075075075e-06,
+      "loss": 0.3398,
+      "step": 64860
+    },
+    {
+      "epoch": 194.8,
+      "grad_norm": 12.734922409057617,
+      "learning_rate": 3.506506506506507e-06,
+      "loss": 0.3466,
+      "step": 64870
+    },
+    {
+      "epoch": 194.83,
+      "grad_norm": 14.982222557067871,
+      "learning_rate": 3.505505505505506e-06,
+      "loss": 0.3927,
+      "step": 64880
+    },
+    {
+      "epoch": 194.86,
+      "grad_norm": 30.67961311340332,
+      "learning_rate": 3.5045045045045046e-06,
+      "loss": 0.4084,
+      "step": 64890
+    },
+    {
+      "epoch": 194.89,
+      "grad_norm": 25.25643539428711,
+      "learning_rate": 3.5035035035035036e-06,
+      "loss": 0.3563,
+      "step": 64900
+    },
+    {
+      "epoch": 194.92,
+      "grad_norm": 12.62600326538086,
+      "learning_rate": 3.5025025025025026e-06,
+      "loss": 0.3742,
+      "step": 64910
+    },
+    {
+      "epoch": 194.95,
+      "grad_norm": 10.575051307678223,
+      "learning_rate": 3.501501501501502e-06,
+      "loss": 0.3273,
+      "step": 64920
+    },
+    {
+      "epoch": 194.98,
+      "grad_norm": 14.568055152893066,
+      "learning_rate": 3.500500500500501e-06,
+      "loss": 0.3722,
+      "step": 64930
+    },
+    {
+      "epoch": 195.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5318506956100464,
+      "eval_runtime": 12.4635,
+      "eval_samples_per_second": 802.344,
+      "eval_steps_per_second": 3.209,
+      "step": 64935
+    },
+    {
+      "epoch": 195.02,
+      "grad_norm": 12.320499420166016,
+      "learning_rate": 3.4994994994994997e-06,
+      "loss": 0.323,
+      "step": 64940
+    },
+    {
+      "epoch": 195.05,
+      "grad_norm": 14.970925331115723,
+      "learning_rate": 3.4984984984984987e-06,
+      "loss": 0.3307,
+      "step": 64950
+    },
+    {
+      "epoch": 195.08,
+      "grad_norm": 12.925833702087402,
+      "learning_rate": 3.4974974974974977e-06,
+      "loss": 0.3986,
+      "step": 64960
+    },
+    {
+      "epoch": 195.11,
+      "grad_norm": 13.335636138916016,
+      "learning_rate": 3.496496496496497e-06,
+      "loss": 0.352,
+      "step": 64970
+    },
+    {
+      "epoch": 195.14,
+      "grad_norm": 13.544027328491211,
+      "learning_rate": 3.4954954954954957e-06,
+      "loss": 0.3333,
+      "step": 64980
+    },
+    {
+      "epoch": 195.17,
+      "grad_norm": 26.71274757385254,
+      "learning_rate": 3.4944944944944948e-06,
+      "loss": 0.3691,
+      "step": 64990
+    },
+    {
+      "epoch": 195.2,
+      "grad_norm": 18.05646324157715,
+      "learning_rate": 3.4934934934934938e-06,
+      "loss": 0.3575,
+      "step": 65000
+    },
+    {
+      "epoch": 195.23,
+      "grad_norm": 13.153338432312012,
+      "learning_rate": 3.492492492492493e-06,
+      "loss": 0.3391,
+      "step": 65010
+    },
+    {
+      "epoch": 195.26,
+      "grad_norm": 12.854308128356934,
+      "learning_rate": 3.4914914914914914e-06,
+      "loss": 0.4012,
+      "step": 65020
+    },
+    {
+      "epoch": 195.29,
+      "grad_norm": 14.447165489196777,
+      "learning_rate": 3.490490490490491e-06,
+      "loss": 0.3832,
+      "step": 65030
+    },
+    {
+      "epoch": 195.32,
+      "grad_norm": 15.400614738464355,
+      "learning_rate": 3.48948948948949e-06,
+      "loss": 0.3307,
+      "step": 65040
+    },
+    {
+      "epoch": 195.35,
+      "grad_norm": 17.269357681274414,
+      "learning_rate": 3.488488488488489e-06,
+      "loss": 0.3888,
+      "step": 65050
+    },
+    {
+      "epoch": 195.38,
+      "grad_norm": 15.042293548583984,
+      "learning_rate": 3.4874874874874875e-06,
+      "loss": 0.3049,
+      "step": 65060
+    },
+    {
+      "epoch": 195.41,
+      "grad_norm": 29.293184280395508,
+      "learning_rate": 3.4864864864864865e-06,
+      "loss": 0.3871,
+      "step": 65070
+    },
+    {
+      "epoch": 195.44,
+      "grad_norm": 18.180734634399414,
+      "learning_rate": 3.485485485485486e-06,
+      "loss": 0.3579,
+      "step": 65080
+    },
+    {
+      "epoch": 195.47,
+      "grad_norm": 12.59367847442627,
+      "learning_rate": 3.484484484484485e-06,
+      "loss": 0.3574,
+      "step": 65090
+    },
+    {
+      "epoch": 195.5,
+      "grad_norm": 14.40528392791748,
+      "learning_rate": 3.4834834834834835e-06,
+      "loss": 0.3796,
+      "step": 65100
+    },
+    {
+      "epoch": 195.53,
+      "grad_norm": 23.14837646484375,
+      "learning_rate": 3.4824824824824826e-06,
+      "loss": 0.4094,
+      "step": 65110
+    },
+    {
+      "epoch": 195.56,
+      "grad_norm": 13.695741653442383,
+      "learning_rate": 3.481481481481482e-06,
+      "loss": 0.3286,
+      "step": 65120
+    },
+    {
+      "epoch": 195.59,
+      "grad_norm": 9.257047653198242,
+      "learning_rate": 3.480480480480481e-06,
+      "loss": 0.3333,
+      "step": 65130
+    },
+    {
+      "epoch": 195.62,
+      "grad_norm": 20.210844039916992,
+      "learning_rate": 3.4794794794794796e-06,
+      "loss": 0.3719,
+      "step": 65140
+    },
+    {
+      "epoch": 195.65,
+      "grad_norm": 13.78529167175293,
+      "learning_rate": 3.4784784784784786e-06,
+      "loss": 0.3359,
+      "step": 65150
+    },
+    {
+      "epoch": 195.68,
+      "grad_norm": 18.200393676757812,
+      "learning_rate": 3.4774774774774776e-06,
+      "loss": 0.3662,
+      "step": 65160
+    },
+    {
+      "epoch": 195.71,
+      "grad_norm": 20.060314178466797,
+      "learning_rate": 3.476476476476477e-06,
+      "loss": 0.3891,
+      "step": 65170
+    },
+    {
+      "epoch": 195.74,
+      "grad_norm": 9.10004711151123,
+      "learning_rate": 3.475475475475476e-06,
+      "loss": 0.357,
+      "step": 65180
+    },
+    {
+      "epoch": 195.77,
+      "grad_norm": 17.44548797607422,
+      "learning_rate": 3.4744744744744747e-06,
+      "loss": 0.3542,
+      "step": 65190
+    },
+    {
+      "epoch": 195.8,
+      "grad_norm": 17.05437469482422,
+      "learning_rate": 3.4734734734734737e-06,
+      "loss": 0.3754,
+      "step": 65200
+    },
+    {
+      "epoch": 195.83,
+      "grad_norm": 14.176117897033691,
+      "learning_rate": 3.4724724724724727e-06,
+      "loss": 0.3438,
+      "step": 65210
+    },
+    {
+      "epoch": 195.86,
+      "grad_norm": 14.172228813171387,
+      "learning_rate": 3.471471471471472e-06,
+      "loss": 0.3476,
+      "step": 65220
+    },
+    {
+      "epoch": 195.89,
+      "grad_norm": 14.571975708007812,
+      "learning_rate": 3.4704704704704708e-06,
+      "loss": 0.3264,
+      "step": 65230
+    },
+    {
+      "epoch": 195.92,
+      "grad_norm": 15.168741226196289,
+      "learning_rate": 3.46946946946947e-06,
+      "loss": 0.3541,
+      "step": 65240
+    },
+    {
+      "epoch": 195.95,
+      "grad_norm": 13.87440013885498,
+      "learning_rate": 3.468468468468469e-06,
+      "loss": 0.3087,
+      "step": 65250
+    },
+    {
+      "epoch": 195.98,
+      "grad_norm": 18.606000900268555,
+      "learning_rate": 3.467467467467468e-06,
+      "loss": 0.3525,
+      "step": 65260
+    },
+    {
+      "epoch": 196.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.5347244739532471,
+      "eval_runtime": 12.8907,
+      "eval_samples_per_second": 775.754,
+      "eval_steps_per_second": 3.103,
+      "step": 65268
+    },
+    {
+      "epoch": 196.01,
+      "grad_norm": 19.92584228515625,
+      "learning_rate": 3.4664664664664664e-06,
+      "loss": 0.3358,
+      "step": 65270
+    },
+    {
+      "epoch": 196.04,
+      "grad_norm": 19.191726684570312,
+      "learning_rate": 3.465465465465466e-06,
+      "loss": 0.3967,
+      "step": 65280
+    },
+    {
+      "epoch": 196.07,
+      "grad_norm": 18.884075164794922,
+      "learning_rate": 3.464464464464465e-06,
+      "loss": 0.3686,
+      "step": 65290
+    },
+    {
+      "epoch": 196.1,
+      "grad_norm": 18.111738204956055,
+      "learning_rate": 3.463463463463464e-06,
+      "loss": 0.4006,
+      "step": 65300
+    },
+    {
+      "epoch": 196.13,
+      "grad_norm": 16.535263061523438,
+      "learning_rate": 3.4624624624624625e-06,
+      "loss": 0.3407,
+      "step": 65310
+    },
+    {
+      "epoch": 196.16,
+      "grad_norm": 17.6393985748291,
+      "learning_rate": 3.4614614614614615e-06,
+      "loss": 0.4143,
+      "step": 65320
+    },
+    {
+      "epoch": 196.19,
+      "grad_norm": 11.103723526000977,
+      "learning_rate": 3.460460460460461e-06,
+      "loss": 0.4035,
+      "step": 65330
+    },
+    {
+      "epoch": 196.22,
+      "grad_norm": 13.215161323547363,
+      "learning_rate": 3.45945945945946e-06,
+      "loss": 0.4065,
+      "step": 65340
+    },
+    {
+      "epoch": 196.25,
+      "grad_norm": 17.009273529052734,
+      "learning_rate": 3.4584584584584586e-06,
+      "loss": 0.3146,
+      "step": 65350
+    },
+    {
+      "epoch": 196.28,
+      "grad_norm": 12.088274002075195,
+      "learning_rate": 3.4574574574574576e-06,
+      "loss": 0.3247,
+      "step": 65360
+    },
+    {
+      "epoch": 196.31,
+      "grad_norm": 15.409132957458496,
+      "learning_rate": 3.4564564564564566e-06,
+      "loss": 0.3285,
+      "step": 65370
+    },
+    {
+      "epoch": 196.34,
+      "grad_norm": 15.337315559387207,
+      "learning_rate": 3.455455455455456e-06,
+      "loss": 0.322,
+      "step": 65380
+    },
+    {
+      "epoch": 196.37,
+      "grad_norm": 15.986523628234863,
+      "learning_rate": 3.4544544544544546e-06,
+      "loss": 0.3484,
+      "step": 65390
+    },
+    {
+      "epoch": 196.4,
+      "grad_norm": 17.414743423461914,
+      "learning_rate": 3.4534534534534537e-06,
+      "loss": 0.3718,
+      "step": 65400
+    },
+    {
+      "epoch": 196.43,
+      "grad_norm": 15.904192924499512,
+      "learning_rate": 3.4524524524524527e-06,
+      "loss": 0.3307,
+      "step": 65410
+    },
+    {
+      "epoch": 196.46,
+      "grad_norm": 13.843870162963867,
+      "learning_rate": 3.4514514514514517e-06,
+      "loss": 0.359,
+      "step": 65420
+    },
+    {
+      "epoch": 196.49,
+      "grad_norm": 18.749446868896484,
+      "learning_rate": 3.4504504504504503e-06,
+      "loss": 0.3672,
+      "step": 65430
+    },
+    {
+      "epoch": 196.52,
+      "grad_norm": 19.063426971435547,
+      "learning_rate": 3.4494494494494497e-06,
+      "loss": 0.3803,
+      "step": 65440
+    },
+    {
+      "epoch": 196.55,
+      "grad_norm": 27.517181396484375,
+      "learning_rate": 3.4484484484484488e-06,
+      "loss": 0.3757,
+      "step": 65450
+    },
+    {
+      "epoch": 196.58,
+      "grad_norm": 13.47083854675293,
+      "learning_rate": 3.4474474474474478e-06,
+      "loss": 0.3781,
+      "step": 65460
+    },
+    {
+      "epoch": 196.61,
+      "grad_norm": 11.18587875366211,
+      "learning_rate": 3.446446446446447e-06,
+      "loss": 0.3519,
+      "step": 65470
+    },
+    {
+      "epoch": 196.64,
+      "grad_norm": 14.283446311950684,
+      "learning_rate": 3.4454454454454454e-06,
+      "loss": 0.3067,
+      "step": 65480
+    },
+    {
+      "epoch": 196.67,
+      "grad_norm": 17.500286102294922,
+      "learning_rate": 3.444444444444445e-06,
+      "loss": 0.3293,
+      "step": 65490
+    },
+    {
+      "epoch": 196.7,
+      "grad_norm": 24.184053421020508,
+      "learning_rate": 3.443443443443444e-06,
+      "loss": 0.3779,
+      "step": 65500
+    },
+    {
+      "epoch": 196.73,
+      "grad_norm": 14.31028938293457,
+      "learning_rate": 3.442442442442443e-06,
+      "loss": 0.3755,
+      "step": 65510
+    },
+    {
+      "epoch": 196.76,
+      "grad_norm": 16.857194900512695,
+      "learning_rate": 3.4414414414414415e-06,
+      "loss": 0.3741,
+      "step": 65520
+    },
+    {
+      "epoch": 196.79,
+      "grad_norm": 11.496053695678711,
+      "learning_rate": 3.4404404404404405e-06,
+      "loss": 0.3552,
+      "step": 65530
+    },
+    {
+      "epoch": 196.82,
+      "grad_norm": 15.868075370788574,
+      "learning_rate": 3.43943943943944e-06,
+      "loss": 0.3546,
+      "step": 65540
+    },
+    {
+      "epoch": 196.85,
+      "grad_norm": 21.518115997314453,
+      "learning_rate": 3.438438438438439e-06,
+      "loss": 0.335,
+      "step": 65550
+    },
+    {
+      "epoch": 196.88,
+      "grad_norm": 23.70357322692871,
+      "learning_rate": 3.4374374374374375e-06,
+      "loss": 0.3872,
+      "step": 65560
+    },
+    {
+      "epoch": 196.91,
+      "grad_norm": 13.535188674926758,
+      "learning_rate": 3.4364364364364366e-06,
+      "loss": 0.3792,
+      "step": 65570
+    },
+    {
+      "epoch": 196.94,
+      "grad_norm": 17.806114196777344,
+      "learning_rate": 3.435435435435436e-06,
+      "loss": 0.3617,
+      "step": 65580
+    },
+    {
+      "epoch": 196.97,
+      "grad_norm": 16.02191162109375,
+      "learning_rate": 3.434434434434435e-06,
+      "loss": 0.337,
+      "step": 65590
+    },
+    {
+      "epoch": 197.0,
+      "grad_norm": 17.437562942504883,
+      "learning_rate": 3.4334334334334336e-06,
+      "loss": 0.3561,
+      "step": 65600
+    },
+    {
+      "epoch": 197.0,
+      "eval_accuracy": 0.8629,
+      "eval_loss": 0.5406749844551086,
+      "eval_runtime": 12.8434,
+      "eval_samples_per_second": 778.612,
+      "eval_steps_per_second": 3.114,
+      "step": 65601
+    },
+    {
+      "epoch": 197.03,
+      "grad_norm": 12.99141788482666,
+      "learning_rate": 3.4324324324324326e-06,
+      "loss": 0.3526,
+      "step": 65610
+    },
+    {
+      "epoch": 197.06,
+      "grad_norm": 19.357521057128906,
+      "learning_rate": 3.4314314314314316e-06,
+      "loss": 0.3555,
+      "step": 65620
+    },
+    {
+      "epoch": 197.09,
+      "grad_norm": 14.045784950256348,
+      "learning_rate": 3.430430430430431e-06,
+      "loss": 0.4074,
+      "step": 65630
+    },
+    {
+      "epoch": 197.12,
+      "grad_norm": 18.92548370361328,
+      "learning_rate": 3.4294294294294293e-06,
+      "loss": 0.3184,
+      "step": 65640
+    },
+    {
+      "epoch": 197.15,
+      "grad_norm": 20.427648544311523,
+      "learning_rate": 3.4284284284284287e-06,
+      "loss": 0.409,
+      "step": 65650
+    },
+    {
+      "epoch": 197.18,
+      "grad_norm": 11.95384407043457,
+      "learning_rate": 3.4274274274274277e-06,
+      "loss": 0.3849,
+      "step": 65660
+    },
+    {
+      "epoch": 197.21,
+      "grad_norm": 18.84357261657715,
+      "learning_rate": 3.4264264264264267e-06,
+      "loss": 0.3509,
+      "step": 65670
+    },
+    {
+      "epoch": 197.24,
+      "grad_norm": 15.041975021362305,
+      "learning_rate": 3.4254254254254253e-06,
+      "loss": 0.3122,
+      "step": 65680
+    },
+    {
+      "epoch": 197.27,
+      "grad_norm": 18.78207778930664,
+      "learning_rate": 3.4244244244244248e-06,
+      "loss": 0.4085,
+      "step": 65690
+    },
+    {
+      "epoch": 197.3,
+      "grad_norm": 23.061128616333008,
+      "learning_rate": 3.423423423423424e-06,
+      "loss": 0.3183,
+      "step": 65700
+    },
+    {
+      "epoch": 197.33,
+      "grad_norm": 15.286757469177246,
+      "learning_rate": 3.422422422422423e-06,
+      "loss": 0.3727,
+      "step": 65710
+    },
+    {
+      "epoch": 197.36,
+      "grad_norm": 14.683562278747559,
+      "learning_rate": 3.421421421421422e-06,
+      "loss": 0.4476,
+      "step": 65720
+    },
+    {
+      "epoch": 197.39,
+      "grad_norm": 11.786398887634277,
+      "learning_rate": 3.4204204204204204e-06,
+      "loss": 0.353,
+      "step": 65730
+    },
+    {
+      "epoch": 197.42,
+      "grad_norm": 12.613759994506836,
+      "learning_rate": 3.41941941941942e-06,
+      "loss": 0.3476,
+      "step": 65740
+    },
+    {
+      "epoch": 197.45,
+      "grad_norm": 18.94495391845703,
+      "learning_rate": 3.418418418418419e-06,
+      "loss": 0.346,
+      "step": 65750
+    },
+    {
+      "epoch": 197.48,
+      "grad_norm": 9.87179946899414,
+      "learning_rate": 3.417417417417418e-06,
+      "loss": 0.362,
+      "step": 65760
+    },
+    {
+      "epoch": 197.51,
+      "grad_norm": 18.62107276916504,
+      "learning_rate": 3.4164164164164165e-06,
+      "loss": 0.3746,
+      "step": 65770
+    },
+    {
+      "epoch": 197.54,
+      "grad_norm": 16.964862823486328,
+      "learning_rate": 3.4154154154154155e-06,
+      "loss": 0.3323,
+      "step": 65780
+    },
+    {
+      "epoch": 197.57,
+      "grad_norm": 17.735097885131836,
+      "learning_rate": 3.414414414414415e-06,
+      "loss": 0.3893,
+      "step": 65790
+    },
+    {
+      "epoch": 197.6,
+      "grad_norm": 17.71319580078125,
+      "learning_rate": 3.413413413413414e-06,
+      "loss": 0.3546,
+      "step": 65800
+    },
+    {
+      "epoch": 197.63,
+      "grad_norm": 19.584312438964844,
+      "learning_rate": 3.4124124124124126e-06,
+      "loss": 0.3304,
+      "step": 65810
+    },
+    {
+      "epoch": 197.66,
+      "grad_norm": 13.464262962341309,
+      "learning_rate": 3.4114114114114116e-06,
+      "loss": 0.3938,
+      "step": 65820
+    },
+    {
+      "epoch": 197.69,
+      "grad_norm": 16.874616622924805,
+      "learning_rate": 3.4104104104104106e-06,
+      "loss": 0.3434,
+      "step": 65830
+    },
+    {
+      "epoch": 197.72,
+      "grad_norm": 19.323930740356445,
+      "learning_rate": 3.40940940940941e-06,
+      "loss": 0.38,
+      "step": 65840
+    },
+    {
+      "epoch": 197.75,
+      "grad_norm": 10.324000358581543,
+      "learning_rate": 3.4084084084084086e-06,
+      "loss": 0.3914,
+      "step": 65850
+    },
+    {
+      "epoch": 197.78,
+      "grad_norm": 12.550322532653809,
+      "learning_rate": 3.4074074074074077e-06,
+      "loss": 0.3579,
+      "step": 65860
+    },
+    {
+      "epoch": 197.81,
+      "grad_norm": 21.8094482421875,
+      "learning_rate": 3.4064064064064067e-06,
+      "loss": 0.3174,
+      "step": 65870
+    },
+    {
+      "epoch": 197.84,
+      "grad_norm": 23.97180938720703,
+      "learning_rate": 3.4054054054054057e-06,
+      "loss": 0.3598,
+      "step": 65880
+    },
+    {
+      "epoch": 197.87,
+      "grad_norm": 9.735006332397461,
+      "learning_rate": 3.4044044044044043e-06,
+      "loss": 0.3313,
+      "step": 65890
+    },
+    {
+      "epoch": 197.9,
+      "grad_norm": 16.031978607177734,
+      "learning_rate": 3.4034034034034037e-06,
+      "loss": 0.3784,
+      "step": 65900
+    },
+    {
+      "epoch": 197.93,
+      "grad_norm": 12.928275108337402,
+      "learning_rate": 3.4024024024024028e-06,
+      "loss": 0.3473,
+      "step": 65910
+    },
+    {
+      "epoch": 197.96,
+      "grad_norm": 13.23296070098877,
+      "learning_rate": 3.4014014014014018e-06,
+      "loss": 0.3429,
+      "step": 65920
+    },
+    {
+      "epoch": 197.99,
+      "grad_norm": 20.52837371826172,
+      "learning_rate": 3.4004004004004004e-06,
+      "loss": 0.3946,
+      "step": 65930
+    },
+    {
+      "epoch": 198.0,
+      "eval_accuracy": 0.8643,
+      "eval_loss": 0.536087691783905,
+      "eval_runtime": 12.8001,
+      "eval_samples_per_second": 781.245,
+      "eval_steps_per_second": 3.125,
+      "step": 65934
+    },
+    {
+      "epoch": 198.02,
+      "grad_norm": 12.003120422363281,
+      "learning_rate": 3.3993993993993994e-06,
+      "loss": 0.3546,
+      "step": 65940
+    },
+    {
+      "epoch": 198.05,
+      "grad_norm": 16.05032730102539,
+      "learning_rate": 3.398398398398399e-06,
+      "loss": 0.3051,
+      "step": 65950
+    },
+    {
+      "epoch": 198.08,
+      "grad_norm": 15.878999710083008,
+      "learning_rate": 3.397397397397398e-06,
+      "loss": 0.3538,
+      "step": 65960
+    },
+    {
+      "epoch": 198.11,
+      "grad_norm": 21.26803207397461,
+      "learning_rate": 3.3963963963963964e-06,
+      "loss": 0.333,
+      "step": 65970
+    },
+    {
+      "epoch": 198.14,
+      "grad_norm": 22.021371841430664,
+      "learning_rate": 3.3953953953953955e-06,
+      "loss": 0.3574,
+      "step": 65980
+    },
+    {
+      "epoch": 198.17,
+      "grad_norm": 12.458101272583008,
+      "learning_rate": 3.3943943943943945e-06,
+      "loss": 0.3596,
+      "step": 65990
+    },
+    {
+      "epoch": 198.2,
+      "grad_norm": 17.372177124023438,
+      "learning_rate": 3.393393393393394e-06,
+      "loss": 0.3154,
+      "step": 66000
+    },
+    {
+      "epoch": 198.23,
+      "grad_norm": 12.463178634643555,
+      "learning_rate": 3.392392392392393e-06,
+      "loss": 0.3463,
+      "step": 66010
+    },
+    {
+      "epoch": 198.26,
+      "grad_norm": 16.043014526367188,
+      "learning_rate": 3.3913913913913915e-06,
+      "loss": 0.3431,
+      "step": 66020
+    },
+    {
+      "epoch": 198.29,
+      "grad_norm": 16.979982376098633,
+      "learning_rate": 3.3903903903903905e-06,
+      "loss": 0.3254,
+      "step": 66030
+    },
+    {
+      "epoch": 198.32,
+      "grad_norm": 17.209875106811523,
+      "learning_rate": 3.3893893893893896e-06,
+      "loss": 0.3645,
+      "step": 66040
+    },
+    {
+      "epoch": 198.35,
+      "grad_norm": 11.704986572265625,
+      "learning_rate": 3.388388388388389e-06,
+      "loss": 0.3352,
+      "step": 66050
+    },
+    {
+      "epoch": 198.38,
+      "grad_norm": 15.973367691040039,
+      "learning_rate": 3.3873873873873876e-06,
+      "loss": 0.3724,
+      "step": 66060
+    },
+    {
+      "epoch": 198.41,
+      "grad_norm": 15.066311836242676,
+      "learning_rate": 3.3863863863863866e-06,
+      "loss": 0.3548,
+      "step": 66070
+    },
+    {
+      "epoch": 198.44,
+      "grad_norm": 14.314546585083008,
+      "learning_rate": 3.3853853853853856e-06,
+      "loss": 0.3889,
+      "step": 66080
+    },
+    {
+      "epoch": 198.47,
+      "grad_norm": 14.289697647094727,
+      "learning_rate": 3.384384384384385e-06,
+      "loss": 0.329,
+      "step": 66090
+    },
+    {
+      "epoch": 198.5,
+      "grad_norm": 10.343135833740234,
+      "learning_rate": 3.3833833833833833e-06,
+      "loss": 0.3633,
+      "step": 66100
+    },
+    {
+      "epoch": 198.53,
+      "grad_norm": 15.060704231262207,
+      "learning_rate": 3.3823823823823827e-06,
+      "loss": 0.3484,
+      "step": 66110
+    },
+    {
+      "epoch": 198.56,
+      "grad_norm": 19.042139053344727,
+      "learning_rate": 3.3813813813813817e-06,
+      "loss": 0.3833,
+      "step": 66120
+    },
+    {
+      "epoch": 198.59,
+      "grad_norm": 14.288841247558594,
+      "learning_rate": 3.3803803803803807e-06,
+      "loss": 0.3714,
+      "step": 66130
+    },
+    {
+      "epoch": 198.62,
+      "grad_norm": 21.89864730834961,
+      "learning_rate": 3.3793793793793793e-06,
+      "loss": 0.3468,
+      "step": 66140
+    },
+    {
+      "epoch": 198.65,
+      "grad_norm": 15.881699562072754,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.3182,
+      "step": 66150
+    },
+    {
+      "epoch": 198.68,
+      "grad_norm": 12.003392219543457,
+      "learning_rate": 3.3773773773773778e-06,
+      "loss": 0.374,
+      "step": 66160
+    },
+    {
+      "epoch": 198.71,
+      "grad_norm": 23.99654197692871,
+      "learning_rate": 3.376376376376377e-06,
+      "loss": 0.3402,
+      "step": 66170
+    },
+    {
+      "epoch": 198.74,
+      "grad_norm": 16.293447494506836,
+      "learning_rate": 3.3753753753753754e-06,
+      "loss": 0.3701,
+      "step": 66180
+    },
+    {
+      "epoch": 198.77,
+      "grad_norm": 12.801802635192871,
+      "learning_rate": 3.3743743743743744e-06,
+      "loss": 0.3542,
+      "step": 66190
+    },
+    {
+      "epoch": 198.8,
+      "grad_norm": 13.767782211303711,
+      "learning_rate": 3.373373373373374e-06,
+      "loss": 0.3225,
+      "step": 66200
+    },
+    {
+      "epoch": 198.83,
+      "grad_norm": 21.725799560546875,
+      "learning_rate": 3.372372372372373e-06,
+      "loss": 0.4292,
+      "step": 66210
+    },
+    {
+      "epoch": 198.86,
+      "grad_norm": 16.535930633544922,
+      "learning_rate": 3.3713713713713715e-06,
+      "loss": 0.338,
+      "step": 66220
+    },
+    {
+      "epoch": 198.89,
+      "grad_norm": 14.223259925842285,
+      "learning_rate": 3.3703703703703705e-06,
+      "loss": 0.3326,
+      "step": 66230
+    },
+    {
+      "epoch": 198.92,
+      "grad_norm": 12.64869499206543,
+      "learning_rate": 3.3693693693693695e-06,
+      "loss": 0.3187,
+      "step": 66240
+    },
+    {
+      "epoch": 198.95,
+      "grad_norm": 13.60538387298584,
+      "learning_rate": 3.368368368368369e-06,
+      "loss": 0.3746,
+      "step": 66250
+    },
+    {
+      "epoch": 198.98,
+      "grad_norm": 27.915176391601562,
+      "learning_rate": 3.367367367367368e-06,
+      "loss": 0.3768,
+      "step": 66260
+    },
+    {
+      "epoch": 199.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.5386967062950134,
+      "eval_runtime": 12.7842,
+      "eval_samples_per_second": 782.217,
+      "eval_steps_per_second": 3.129,
+      "step": 66267
+    },
+    {
+      "epoch": 199.01,
+      "grad_norm": 19.735395431518555,
+      "learning_rate": 3.3663663663663666e-06,
+      "loss": 0.3276,
+      "step": 66270
+    },
+    {
+      "epoch": 199.04,
+      "grad_norm": 17.286602020263672,
+      "learning_rate": 3.3653653653653656e-06,
+      "loss": 0.3441,
+      "step": 66280
+    },
+    {
+      "epoch": 199.07,
+      "grad_norm": 10.946517944335938,
+      "learning_rate": 3.3643643643643646e-06,
+      "loss": 0.3454,
+      "step": 66290
+    },
+    {
+      "epoch": 199.1,
+      "grad_norm": 22.02970314025879,
+      "learning_rate": 3.363363363363364e-06,
+      "loss": 0.3773,
+      "step": 66300
+    },
+    {
+      "epoch": 199.13,
+      "grad_norm": 14.984495162963867,
+      "learning_rate": 3.3623623623623626e-06,
+      "loss": 0.3277,
+      "step": 66310
+    },
+    {
+      "epoch": 199.16,
+      "grad_norm": 14.183544158935547,
+      "learning_rate": 3.3613613613613617e-06,
+      "loss": 0.3049,
+      "step": 66320
+    },
+    {
+      "epoch": 199.19,
+      "grad_norm": 23.41092300415039,
+      "learning_rate": 3.3603603603603607e-06,
+      "loss": 0.341,
+      "step": 66330
+    },
+    {
+      "epoch": 199.22,
+      "grad_norm": 14.43628978729248,
+      "learning_rate": 3.3593593593593597e-06,
+      "loss": 0.3972,
+      "step": 66340
+    },
+    {
+      "epoch": 199.25,
+      "grad_norm": 13.764900207519531,
+      "learning_rate": 3.3583583583583583e-06,
+      "loss": 0.3727,
+      "step": 66350
+    },
+    {
+      "epoch": 199.28,
+      "grad_norm": 17.670719146728516,
+      "learning_rate": 3.3573573573573577e-06,
+      "loss": 0.3694,
+      "step": 66360
+    },
+    {
+      "epoch": 199.31,
+      "grad_norm": 17.933704376220703,
+      "learning_rate": 3.3563563563563567e-06,
+      "loss": 0.3237,
+      "step": 66370
+    },
+    {
+      "epoch": 199.34,
+      "grad_norm": 13.561380386352539,
+      "learning_rate": 3.3553553553553558e-06,
+      "loss": 0.361,
+      "step": 66380
+    },
+    {
+      "epoch": 199.37,
+      "grad_norm": 17.0720272064209,
+      "learning_rate": 3.3543543543543544e-06,
+      "loss": 0.3572,
+      "step": 66390
+    },
+    {
+      "epoch": 199.4,
+      "grad_norm": 15.703876495361328,
+      "learning_rate": 3.3533533533533534e-06,
+      "loss": 0.3319,
+      "step": 66400
+    },
+    {
+      "epoch": 199.43,
+      "grad_norm": 11.141468048095703,
+      "learning_rate": 3.352352352352353e-06,
+      "loss": 0.3794,
+      "step": 66410
+    },
+    {
+      "epoch": 199.46,
+      "grad_norm": 15.869293212890625,
+      "learning_rate": 3.351351351351352e-06,
+      "loss": 0.3425,
+      "step": 66420
+    },
+    {
+      "epoch": 199.49,
+      "grad_norm": 13.451416969299316,
+      "learning_rate": 3.3503503503503504e-06,
+      "loss": 0.4139,
+      "step": 66430
+    },
+    {
+      "epoch": 199.52,
+      "grad_norm": 11.843367576599121,
+      "learning_rate": 3.3493493493493495e-06,
+      "loss": 0.3812,
+      "step": 66440
+    },
+    {
+      "epoch": 199.55,
+      "grad_norm": 17.737953186035156,
+      "learning_rate": 3.3483483483483485e-06,
+      "loss": 0.3998,
+      "step": 66450
+    },
+    {
+      "epoch": 199.58,
+      "grad_norm": 23.859127044677734,
+      "learning_rate": 3.347347347347348e-06,
+      "loss": 0.3419,
+      "step": 66460
+    },
+    {
+      "epoch": 199.61,
+      "grad_norm": 28.04448127746582,
+      "learning_rate": 3.3463463463463465e-06,
+      "loss": 0.3543,
+      "step": 66470
+    },
+    {
+      "epoch": 199.64,
+      "grad_norm": 15.31539249420166,
+      "learning_rate": 3.3453453453453455e-06,
+      "loss": 0.3437,
+      "step": 66480
+    },
+    {
+      "epoch": 199.67,
+      "grad_norm": 15.551780700683594,
+      "learning_rate": 3.3443443443443445e-06,
+      "loss": 0.34,
+      "step": 66490
+    },
+    {
+      "epoch": 199.7,
+      "grad_norm": 10.951482772827148,
+      "learning_rate": 3.3433433433433436e-06,
+      "loss": 0.3108,
+      "step": 66500
+    },
+    {
+      "epoch": 199.73,
+      "grad_norm": 12.45376968383789,
+      "learning_rate": 3.342342342342342e-06,
+      "loss": 0.3308,
+      "step": 66510
+    },
+    {
+      "epoch": 199.76,
+      "grad_norm": 19.08757209777832,
+      "learning_rate": 3.3413413413413416e-06,
+      "loss": 0.3424,
+      "step": 66520
+    },
+    {
+      "epoch": 199.79,
+      "grad_norm": 19.29985809326172,
+      "learning_rate": 3.3403403403403406e-06,
+      "loss": 0.3954,
+      "step": 66530
+    },
+    {
+      "epoch": 199.82,
+      "grad_norm": 17.144332885742188,
+      "learning_rate": 3.3393393393393396e-06,
+      "loss": 0.3355,
+      "step": 66540
+    },
+    {
+      "epoch": 199.85,
+      "grad_norm": 12.53018856048584,
+      "learning_rate": 3.338338338338339e-06,
+      "loss": 0.3311,
+      "step": 66550
+    },
+    {
+      "epoch": 199.88,
+      "grad_norm": 17.47453498840332,
+      "learning_rate": 3.3373373373373373e-06,
+      "loss": 0.41,
+      "step": 66560
+    },
+    {
+      "epoch": 199.91,
+      "grad_norm": 16.27352523803711,
+      "learning_rate": 3.3363363363363367e-06,
+      "loss": 0.3725,
+      "step": 66570
+    },
+    {
+      "epoch": 199.94,
+      "grad_norm": 26.66290283203125,
+      "learning_rate": 3.3353353353353357e-06,
+      "loss": 0.3729,
+      "step": 66580
+    },
+    {
+      "epoch": 199.97,
+      "grad_norm": 17.14523696899414,
+      "learning_rate": 3.3343343343343347e-06,
+      "loss": 0.3565,
+      "step": 66590
+    },
+    {
+      "epoch": 200.0,
+      "grad_norm": 0.6842318773269653,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3328,
+      "step": 66600
+    },
+    {
+      "epoch": 200.0,
+      "eval_accuracy": 0.8656,
+      "eval_loss": 0.5325269103050232,
+      "eval_runtime": 12.7961,
+      "eval_samples_per_second": 781.491,
+      "eval_steps_per_second": 3.126,
+      "step": 66600
+    },
+    {
+      "epoch": 200.03,
+      "grad_norm": 15.133691787719727,
+      "learning_rate": 3.3323323323323323e-06,
+      "loss": 0.3501,
+      "step": 66610
+    },
+    {
+      "epoch": 200.06,
+      "grad_norm": 17.224506378173828,
+      "learning_rate": 3.3313313313313318e-06,
+      "loss": 0.3641,
+      "step": 66620
+    },
+    {
+      "epoch": 200.09,
+      "grad_norm": 14.254813194274902,
+      "learning_rate": 3.330330330330331e-06,
+      "loss": 0.3591,
+      "step": 66630
+    },
+    {
+      "epoch": 200.12,
+      "grad_norm": 20.212072372436523,
+      "learning_rate": 3.3293293293293294e-06,
+      "loss": 0.3556,
+      "step": 66640
+    },
+    {
+      "epoch": 200.15,
+      "grad_norm": 21.4140682220459,
+      "learning_rate": 3.3283283283283284e-06,
+      "loss": 0.3466,
+      "step": 66650
+    },
+    {
+      "epoch": 200.18,
+      "grad_norm": 15.238445281982422,
+      "learning_rate": 3.327327327327328e-06,
+      "loss": 0.3729,
+      "step": 66660
+    },
+    {
+      "epoch": 200.21,
+      "grad_norm": 11.798453330993652,
+      "learning_rate": 3.326326326326327e-06,
+      "loss": 0.3802,
+      "step": 66670
+    },
+    {
+      "epoch": 200.24,
+      "grad_norm": 17.47968864440918,
+      "learning_rate": 3.3253253253253255e-06,
+      "loss": 0.3231,
+      "step": 66680
+    },
+    {
+      "epoch": 200.27,
+      "grad_norm": 17.861671447753906,
+      "learning_rate": 3.3243243243243245e-06,
+      "loss": 0.3423,
+      "step": 66690
+    },
+    {
+      "epoch": 200.3,
+      "grad_norm": 23.21867561340332,
+      "learning_rate": 3.3233233233233235e-06,
+      "loss": 0.3874,
+      "step": 66700
+    },
+    {
+      "epoch": 200.33,
+      "grad_norm": 23.803476333618164,
+      "learning_rate": 3.322322322322323e-06,
+      "loss": 0.3789,
+      "step": 66710
+    },
+    {
+      "epoch": 200.36,
+      "grad_norm": 13.263434410095215,
+      "learning_rate": 3.3213213213213215e-06,
+      "loss": 0.3549,
+      "step": 66720
+    },
+    {
+      "epoch": 200.39,
+      "grad_norm": 12.681262016296387,
+      "learning_rate": 3.3203203203203206e-06,
+      "loss": 0.332,
+      "step": 66730
+    },
+    {
+      "epoch": 200.42,
+      "grad_norm": 10.657028198242188,
+      "learning_rate": 3.3193193193193196e-06,
+      "loss": 0.3421,
+      "step": 66740
+    },
+    {
+      "epoch": 200.45,
+      "grad_norm": 18.08957290649414,
+      "learning_rate": 3.3183183183183186e-06,
+      "loss": 0.3502,
+      "step": 66750
+    },
+    {
+      "epoch": 200.48,
+      "grad_norm": 19.387685775756836,
+      "learning_rate": 3.317317317317317e-06,
+      "loss": 0.3082,
+      "step": 66760
+    },
+    {
+      "epoch": 200.51,
+      "grad_norm": 14.200443267822266,
+      "learning_rate": 3.3163163163163166e-06,
+      "loss": 0.3262,
+      "step": 66770
+    },
+    {
+      "epoch": 200.54,
+      "grad_norm": 13.051509857177734,
+      "learning_rate": 3.3153153153153157e-06,
+      "loss": 0.3214,
+      "step": 66780
+    },
+    {
+      "epoch": 200.57,
+      "grad_norm": 15.967508316040039,
+      "learning_rate": 3.3143143143143147e-06,
+      "loss": 0.3293,
+      "step": 66790
+    },
+    {
+      "epoch": 200.6,
+      "grad_norm": 16.033424377441406,
+      "learning_rate": 3.3133133133133137e-06,
+      "loss": 0.3578,
+      "step": 66800
+    },
+    {
+      "epoch": 200.63,
+      "grad_norm": 12.742061614990234,
+      "learning_rate": 3.3123123123123123e-06,
+      "loss": 0.332,
+      "step": 66810
+    },
+    {
+      "epoch": 200.66,
+      "grad_norm": 22.767105102539062,
+      "learning_rate": 3.3113113113113117e-06,
+      "loss": 0.3661,
+      "step": 66820
+    },
+    {
+      "epoch": 200.69,
+      "grad_norm": 13.140266418457031,
+      "learning_rate": 3.3103103103103107e-06,
+      "loss": 0.4383,
+      "step": 66830
+    },
+    {
+      "epoch": 200.72,
+      "grad_norm": 22.558027267456055,
+      "learning_rate": 3.3093093093093098e-06,
+      "loss": 0.3853,
+      "step": 66840
+    },
+    {
+      "epoch": 200.75,
+      "grad_norm": 14.777522087097168,
+      "learning_rate": 3.3083083083083084e-06,
+      "loss": 0.3932,
+      "step": 66850
+    },
+    {
+      "epoch": 200.78,
+      "grad_norm": 13.468594551086426,
+      "learning_rate": 3.3073073073073074e-06,
+      "loss": 0.3257,
+      "step": 66860
+    },
+    {
+      "epoch": 200.81,
+      "grad_norm": 19.38922119140625,
+      "learning_rate": 3.306306306306307e-06,
+      "loss": 0.3734,
+      "step": 66870
+    },
+    {
+      "epoch": 200.84,
+      "grad_norm": 14.327119827270508,
+      "learning_rate": 3.305305305305306e-06,
+      "loss": 0.3474,
+      "step": 66880
+    },
+    {
+      "epoch": 200.87,
+      "grad_norm": 16.82010269165039,
+      "learning_rate": 3.3043043043043044e-06,
+      "loss": 0.3408,
+      "step": 66890
+    },
+    {
+      "epoch": 200.9,
+      "grad_norm": 18.912498474121094,
+      "learning_rate": 3.3033033033033035e-06,
+      "loss": 0.3665,
+      "step": 66900
+    },
+    {
+      "epoch": 200.93,
+      "grad_norm": 20.573020935058594,
+      "learning_rate": 3.3023023023023025e-06,
+      "loss": 0.3892,
+      "step": 66910
+    },
+    {
+      "epoch": 200.96,
+      "grad_norm": 14.25319766998291,
+      "learning_rate": 3.301301301301302e-06,
+      "loss": 0.3696,
+      "step": 66920
+    },
+    {
+      "epoch": 200.99,
+      "grad_norm": 12.679616928100586,
+      "learning_rate": 3.3003003003003005e-06,
+      "loss": 0.3418,
+      "step": 66930
+    },
+    {
+      "epoch": 201.0,
+      "eval_accuracy": 0.8676,
+      "eval_loss": 0.530605673789978,
+      "eval_runtime": 12.86,
+      "eval_samples_per_second": 777.603,
+      "eval_steps_per_second": 3.11,
+      "step": 66933
+    },
+    {
+      "epoch": 201.02,
+      "grad_norm": 17.06678581237793,
+      "learning_rate": 3.2992992992992995e-06,
+      "loss": 0.3138,
+      "step": 66940
+    },
+    {
+      "epoch": 201.05,
+      "grad_norm": 12.681791305541992,
+      "learning_rate": 3.2982982982982985e-06,
+      "loss": 0.3762,
+      "step": 66950
+    },
+    {
+      "epoch": 201.08,
+      "grad_norm": 14.155091285705566,
+      "learning_rate": 3.2972972972972976e-06,
+      "loss": 0.3293,
+      "step": 66960
+    },
+    {
+      "epoch": 201.11,
+      "grad_norm": 15.8582763671875,
+      "learning_rate": 3.296296296296296e-06,
+      "loss": 0.3454,
+      "step": 66970
+    },
+    {
+      "epoch": 201.14,
+      "grad_norm": 22.364328384399414,
+      "learning_rate": 3.2952952952952956e-06,
+      "loss": 0.3713,
+      "step": 66980
+    },
+    {
+      "epoch": 201.17,
+      "grad_norm": 13.58890438079834,
+      "learning_rate": 3.2942942942942946e-06,
+      "loss": 0.3495,
+      "step": 66990
+    },
+    {
+      "epoch": 201.2,
+      "grad_norm": 20.498464584350586,
+      "learning_rate": 3.2932932932932936e-06,
+      "loss": 0.351,
+      "step": 67000
+    },
+    {
+      "epoch": 201.23,
+      "grad_norm": 14.51928997039795,
+      "learning_rate": 3.2922922922922922e-06,
+      "loss": 0.3418,
+      "step": 67010
+    },
+    {
+      "epoch": 201.26,
+      "grad_norm": 13.846062660217285,
+      "learning_rate": 3.2912912912912912e-06,
+      "loss": 0.3266,
+      "step": 67020
+    },
+    {
+      "epoch": 201.29,
+      "grad_norm": 9.656928062438965,
+      "learning_rate": 3.2902902902902907e-06,
+      "loss": 0.375,
+      "step": 67030
+    },
+    {
+      "epoch": 201.32,
+      "grad_norm": 19.169897079467773,
+      "learning_rate": 3.2892892892892897e-06,
+      "loss": 0.3344,
+      "step": 67040
+    },
+    {
+      "epoch": 201.35,
+      "grad_norm": 18.013978958129883,
+      "learning_rate": 3.2882882882882887e-06,
+      "loss": 0.3287,
+      "step": 67050
+    },
+    {
+      "epoch": 201.38,
+      "grad_norm": 12.221697807312012,
+      "learning_rate": 3.2872872872872873e-06,
+      "loss": 0.3328,
+      "step": 67060
+    },
+    {
+      "epoch": 201.41,
+      "grad_norm": 16.92691421508789,
+      "learning_rate": 3.2862862862862863e-06,
+      "loss": 0.3065,
+      "step": 67070
+    },
+    {
+      "epoch": 201.44,
+      "grad_norm": 23.58794403076172,
+      "learning_rate": 3.2852852852852858e-06,
+      "loss": 0.3056,
+      "step": 67080
+    },
+    {
+      "epoch": 201.47,
+      "grad_norm": 12.591498374938965,
+      "learning_rate": 3.284284284284285e-06,
+      "loss": 0.3761,
+      "step": 67090
+    },
+    {
+      "epoch": 201.5,
+      "grad_norm": 10.654427528381348,
+      "learning_rate": 3.2832832832832834e-06,
+      "loss": 0.3307,
+      "step": 67100
+    },
+    {
+      "epoch": 201.53,
+      "grad_norm": 12.256685256958008,
+      "learning_rate": 3.2822822822822824e-06,
+      "loss": 0.3523,
+      "step": 67110
+    },
+    {
+      "epoch": 201.56,
+      "grad_norm": 12.532700538635254,
+      "learning_rate": 3.281281281281282e-06,
+      "loss": 0.3115,
+      "step": 67120
+    },
+    {
+      "epoch": 201.59,
+      "grad_norm": 17.75162696838379,
+      "learning_rate": 3.280280280280281e-06,
+      "loss": 0.3564,
+      "step": 67130
+    },
+    {
+      "epoch": 201.62,
+      "grad_norm": 18.37679672241211,
+      "learning_rate": 3.2792792792792795e-06,
+      "loss": 0.3786,
+      "step": 67140
+    },
+    {
+      "epoch": 201.65,
+      "grad_norm": 21.32754898071289,
+      "learning_rate": 3.2782782782782785e-06,
+      "loss": 0.3812,
+      "step": 67150
+    },
+    {
+      "epoch": 201.68,
+      "grad_norm": 12.842366218566895,
+      "learning_rate": 3.2772772772772775e-06,
+      "loss": 0.3129,
+      "step": 67160
+    },
+    {
+      "epoch": 201.71,
+      "grad_norm": 15.541666984558105,
+      "learning_rate": 3.276276276276277e-06,
+      "loss": 0.3989,
+      "step": 67170
+    },
+    {
+      "epoch": 201.74,
+      "grad_norm": 21.269243240356445,
+      "learning_rate": 3.2752752752752755e-06,
+      "loss": 0.3917,
+      "step": 67180
+    },
+    {
+      "epoch": 201.77,
+      "grad_norm": 24.668994903564453,
+      "learning_rate": 3.2742742742742746e-06,
+      "loss": 0.4095,
+      "step": 67190
+    },
+    {
+      "epoch": 201.8,
+      "grad_norm": 15.691615104675293,
+      "learning_rate": 3.2732732732732736e-06,
+      "loss": 0.3943,
+      "step": 67200
+    },
+    {
+      "epoch": 201.83,
+      "grad_norm": 10.370479583740234,
+      "learning_rate": 3.2722722722722726e-06,
+      "loss": 0.3257,
+      "step": 67210
+    },
+    {
+      "epoch": 201.86,
+      "grad_norm": 28.563413619995117,
+      "learning_rate": 3.271271271271271e-06,
+      "loss": 0.356,
+      "step": 67220
+    },
+    {
+      "epoch": 201.89,
+      "grad_norm": 14.545961380004883,
+      "learning_rate": 3.2702702702702706e-06,
+      "loss": 0.3248,
+      "step": 67230
+    },
+    {
+      "epoch": 201.92,
+      "grad_norm": 13.428220748901367,
+      "learning_rate": 3.2692692692692696e-06,
+      "loss": 0.3294,
+      "step": 67240
+    },
+    {
+      "epoch": 201.95,
+      "grad_norm": 17.09244728088379,
+      "learning_rate": 3.2682682682682687e-06,
+      "loss": 0.4094,
+      "step": 67250
+    },
+    {
+      "epoch": 201.98,
+      "grad_norm": 20.891996383666992,
+      "learning_rate": 3.2672672672672673e-06,
+      "loss": 0.3542,
+      "step": 67260
+    },
+    {
+      "epoch": 202.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5320509672164917,
+      "eval_runtime": 12.8129,
+      "eval_samples_per_second": 780.462,
+      "eval_steps_per_second": 3.122,
+      "step": 67266
+    },
+    {
+      "epoch": 202.01,
+      "grad_norm": 15.040534019470215,
+      "learning_rate": 3.2662662662662663e-06,
+      "loss": 0.4792,
+      "step": 67270
+    },
+    {
+      "epoch": 202.04,
+      "grad_norm": 15.534871101379395,
+      "learning_rate": 3.2652652652652657e-06,
+      "loss": 0.3304,
+      "step": 67280
+    },
+    {
+      "epoch": 202.07,
+      "grad_norm": 12.367944717407227,
+      "learning_rate": 3.2642642642642647e-06,
+      "loss": 0.3369,
+      "step": 67290
+    },
+    {
+      "epoch": 202.1,
+      "grad_norm": 11.486591339111328,
+      "learning_rate": 3.2632632632632633e-06,
+      "loss": 0.3801,
+      "step": 67300
+    },
+    {
+      "epoch": 202.13,
+      "grad_norm": 25.76357078552246,
+      "learning_rate": 3.2622622622622624e-06,
+      "loss": 0.3046,
+      "step": 67310
+    },
+    {
+      "epoch": 202.16,
+      "grad_norm": 13.070442199707031,
+      "learning_rate": 3.2612612612612614e-06,
+      "loss": 0.327,
+      "step": 67320
+    },
+    {
+      "epoch": 202.19,
+      "grad_norm": 17.124940872192383,
+      "learning_rate": 3.260260260260261e-06,
+      "loss": 0.4233,
+      "step": 67330
+    },
+    {
+      "epoch": 202.22,
+      "grad_norm": 20.428468704223633,
+      "learning_rate": 3.25925925925926e-06,
+      "loss": 0.3523,
+      "step": 67340
+    },
+    {
+      "epoch": 202.25,
+      "grad_norm": 21.511611938476562,
+      "learning_rate": 3.2582582582582584e-06,
+      "loss": 0.3602,
+      "step": 67350
+    },
+    {
+      "epoch": 202.28,
+      "grad_norm": 17.28934097290039,
+      "learning_rate": 3.2572572572572574e-06,
+      "loss": 0.3304,
+      "step": 67360
+    },
+    {
+      "epoch": 202.31,
+      "grad_norm": 13.798293113708496,
+      "learning_rate": 3.2562562562562565e-06,
+      "loss": 0.3478,
+      "step": 67370
+    },
+    {
+      "epoch": 202.34,
+      "grad_norm": 15.152268409729004,
+      "learning_rate": 3.255255255255256e-06,
+      "loss": 0.3851,
+      "step": 67380
+    },
+    {
+      "epoch": 202.37,
+      "grad_norm": 13.623045921325684,
+      "learning_rate": 3.2542542542542545e-06,
+      "loss": 0.3149,
+      "step": 67390
+    },
+    {
+      "epoch": 202.4,
+      "grad_norm": 17.52507972717285,
+      "learning_rate": 3.2532532532532535e-06,
+      "loss": 0.3248,
+      "step": 67400
+    },
+    {
+      "epoch": 202.43,
+      "grad_norm": 16.104248046875,
+      "learning_rate": 3.2522522522522525e-06,
+      "loss": 0.3459,
+      "step": 67410
+    },
+    {
+      "epoch": 202.46,
+      "grad_norm": 14.086113929748535,
+      "learning_rate": 3.2512512512512516e-06,
+      "loss": 0.3677,
+      "step": 67420
+    },
+    {
+      "epoch": 202.49,
+      "grad_norm": 17.387744903564453,
+      "learning_rate": 3.25025025025025e-06,
+      "loss": 0.3126,
+      "step": 67430
+    },
+    {
+      "epoch": 202.52,
+      "grad_norm": 12.02539348602295,
+      "learning_rate": 3.2492492492492496e-06,
+      "loss": 0.2989,
+      "step": 67440
+    },
+    {
+      "epoch": 202.55,
+      "grad_norm": 13.552552223205566,
+      "learning_rate": 3.2482482482482486e-06,
+      "loss": 0.3012,
+      "step": 67450
+    },
+    {
+      "epoch": 202.58,
+      "grad_norm": 18.374845504760742,
+      "learning_rate": 3.2472472472472476e-06,
+      "loss": 0.3169,
+      "step": 67460
+    },
+    {
+      "epoch": 202.61,
+      "grad_norm": 21.943395614624023,
+      "learning_rate": 3.2462462462462462e-06,
+      "loss": 0.3439,
+      "step": 67470
+    },
+    {
+      "epoch": 202.64,
+      "grad_norm": 15.226884841918945,
+      "learning_rate": 3.2452452452452452e-06,
+      "loss": 0.3454,
+      "step": 67480
+    },
+    {
+      "epoch": 202.67,
+      "grad_norm": 14.208313941955566,
+      "learning_rate": 3.2442442442442447e-06,
+      "loss": 0.3126,
+      "step": 67490
+    },
+    {
+      "epoch": 202.7,
+      "grad_norm": 18.038480758666992,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.347,
+      "step": 67500
+    },
+    {
+      "epoch": 202.73,
+      "grad_norm": 13.097338676452637,
+      "learning_rate": 3.2422422422422423e-06,
+      "loss": 0.3224,
+      "step": 67510
+    },
+    {
+      "epoch": 202.76,
+      "grad_norm": 29.13762855529785,
+      "learning_rate": 3.2412412412412413e-06,
+      "loss": 0.3996,
+      "step": 67520
+    },
+    {
+      "epoch": 202.79,
+      "grad_norm": 19.168672561645508,
+      "learning_rate": 3.2402402402402403e-06,
+      "loss": 0.4259,
+      "step": 67530
+    },
+    {
+      "epoch": 202.82,
+      "grad_norm": 16.77165985107422,
+      "learning_rate": 3.2392392392392398e-06,
+      "loss": 0.3736,
+      "step": 67540
+    },
+    {
+      "epoch": 202.85,
+      "grad_norm": 18.176279067993164,
+      "learning_rate": 3.2382382382382384e-06,
+      "loss": 0.3791,
+      "step": 67550
+    },
+    {
+      "epoch": 202.88,
+      "grad_norm": 15.641382217407227,
+      "learning_rate": 3.2372372372372374e-06,
+      "loss": 0.3374,
+      "step": 67560
+    },
+    {
+      "epoch": 202.91,
+      "grad_norm": 18.018041610717773,
+      "learning_rate": 3.2362362362362364e-06,
+      "loss": 0.3775,
+      "step": 67570
+    },
+    {
+      "epoch": 202.94,
+      "grad_norm": 10.46972370147705,
+      "learning_rate": 3.235235235235236e-06,
+      "loss": 0.3187,
+      "step": 67580
+    },
+    {
+      "epoch": 202.97,
+      "grad_norm": 12.9953031539917,
+      "learning_rate": 3.234234234234235e-06,
+      "loss": 0.3688,
+      "step": 67590
+    },
+    {
+      "epoch": 203.0,
+      "eval_accuracy": 0.8598,
+      "eval_loss": 0.5429918766021729,
+      "eval_runtime": 12.912,
+      "eval_samples_per_second": 774.471,
+      "eval_steps_per_second": 3.098,
+      "step": 67599
+    },
+    {
+      "epoch": 203.0,
+      "grad_norm": 23.347192764282227,
+      "learning_rate": 3.2332332332332335e-06,
+      "loss": 0.4776,
+      "step": 67600
+    },
+    {
+      "epoch": 203.03,
+      "grad_norm": 23.624309539794922,
+      "learning_rate": 3.2322322322322325e-06,
+      "loss": 0.4071,
+      "step": 67610
+    },
+    {
+      "epoch": 203.06,
+      "grad_norm": 16.25799560546875,
+      "learning_rate": 3.2312312312312315e-06,
+      "loss": 0.3296,
+      "step": 67620
+    },
+    {
+      "epoch": 203.09,
+      "grad_norm": 17.69864845275879,
+      "learning_rate": 3.230230230230231e-06,
+      "loss": 0.3344,
+      "step": 67630
+    },
+    {
+      "epoch": 203.12,
+      "grad_norm": 13.574851036071777,
+      "learning_rate": 3.229229229229229e-06,
+      "loss": 0.3711,
+      "step": 67640
+    },
+    {
+      "epoch": 203.15,
+      "grad_norm": 19.36281394958496,
+      "learning_rate": 3.2282282282282286e-06,
+      "loss": 0.3579,
+      "step": 67650
+    },
+    {
+      "epoch": 203.18,
+      "grad_norm": 12.800436973571777,
+      "learning_rate": 3.2272272272272276e-06,
+      "loss": 0.3191,
+      "step": 67660
+    },
+    {
+      "epoch": 203.21,
+      "grad_norm": 21.071374893188477,
+      "learning_rate": 3.2262262262262266e-06,
+      "loss": 0.377,
+      "step": 67670
+    },
+    {
+      "epoch": 203.24,
+      "grad_norm": 19.05280876159668,
+      "learning_rate": 3.225225225225225e-06,
+      "loss": 0.3574,
+      "step": 67680
+    },
+    {
+      "epoch": 203.27,
+      "grad_norm": 15.893171310424805,
+      "learning_rate": 3.2242242242242246e-06,
+      "loss": 0.3314,
+      "step": 67690
+    },
+    {
+      "epoch": 203.3,
+      "grad_norm": 20.597824096679688,
+      "learning_rate": 3.2232232232232236e-06,
+      "loss": 0.3491,
+      "step": 67700
+    },
+    {
+      "epoch": 203.33,
+      "grad_norm": 24.568357467651367,
+      "learning_rate": 3.2222222222222227e-06,
+      "loss": 0.3591,
+      "step": 67710
+    },
+    {
+      "epoch": 203.36,
+      "grad_norm": 14.344891548156738,
+      "learning_rate": 3.2212212212212213e-06,
+      "loss": 0.3626,
+      "step": 67720
+    },
+    {
+      "epoch": 203.39,
+      "grad_norm": 14.527661323547363,
+      "learning_rate": 3.2202202202202203e-06,
+      "loss": 0.3481,
+      "step": 67730
+    },
+    {
+      "epoch": 203.42,
+      "grad_norm": 14.687244415283203,
+      "learning_rate": 3.2192192192192197e-06,
+      "loss": 0.3381,
+      "step": 67740
+    },
+    {
+      "epoch": 203.45,
+      "grad_norm": 37.460086822509766,
+      "learning_rate": 3.2182182182182187e-06,
+      "loss": 0.3953,
+      "step": 67750
+    },
+    {
+      "epoch": 203.48,
+      "grad_norm": 15.797501564025879,
+      "learning_rate": 3.2172172172172173e-06,
+      "loss": 0.3532,
+      "step": 67760
+    },
+    {
+      "epoch": 203.51,
+      "grad_norm": 17.804569244384766,
+      "learning_rate": 3.2162162162162164e-06,
+      "loss": 0.31,
+      "step": 67770
+    },
+    {
+      "epoch": 203.54,
+      "grad_norm": 15.473325729370117,
+      "learning_rate": 3.2152152152152154e-06,
+      "loss": 0.3427,
+      "step": 67780
+    },
+    {
+      "epoch": 203.57,
+      "grad_norm": 11.86602783203125,
+      "learning_rate": 3.214214214214215e-06,
+      "loss": 0.3242,
+      "step": 67790
+    },
+    {
+      "epoch": 203.6,
+      "grad_norm": 17.575504302978516,
+      "learning_rate": 3.2132132132132134e-06,
+      "loss": 0.4139,
+      "step": 67800
+    },
+    {
+      "epoch": 203.63,
+      "grad_norm": 11.399277687072754,
+      "learning_rate": 3.2122122122122124e-06,
+      "loss": 0.3748,
+      "step": 67810
+    },
+    {
+      "epoch": 203.66,
+      "grad_norm": 23.228588104248047,
+      "learning_rate": 3.2112112112112114e-06,
+      "loss": 0.3839,
+      "step": 67820
+    },
+    {
+      "epoch": 203.69,
+      "grad_norm": 21.104352951049805,
+      "learning_rate": 3.2102102102102105e-06,
+      "loss": 0.3613,
+      "step": 67830
+    },
+    {
+      "epoch": 203.72,
+      "grad_norm": 18.415775299072266,
+      "learning_rate": 3.209209209209209e-06,
+      "loss": 0.3527,
+      "step": 67840
+    },
+    {
+      "epoch": 203.75,
+      "grad_norm": 23.890090942382812,
+      "learning_rate": 3.2082082082082085e-06,
+      "loss": 0.3739,
+      "step": 67850
+    },
+    {
+      "epoch": 203.78,
+      "grad_norm": 16.738710403442383,
+      "learning_rate": 3.2072072072072075e-06,
+      "loss": 0.3292,
+      "step": 67860
+    },
+    {
+      "epoch": 203.81,
+      "grad_norm": 20.681047439575195,
+      "learning_rate": 3.2062062062062065e-06,
+      "loss": 0.349,
+      "step": 67870
+    },
+    {
+      "epoch": 203.84,
+      "grad_norm": 20.26121711730957,
+      "learning_rate": 3.2052052052052056e-06,
+      "loss": 0.3389,
+      "step": 67880
+    },
+    {
+      "epoch": 203.87,
+      "grad_norm": 14.120149612426758,
+      "learning_rate": 3.204204204204204e-06,
+      "loss": 0.3707,
+      "step": 67890
+    },
+    {
+      "epoch": 203.9,
+      "grad_norm": 23.198190689086914,
+      "learning_rate": 3.2032032032032036e-06,
+      "loss": 0.3223,
+      "step": 67900
+    },
+    {
+      "epoch": 203.93,
+      "grad_norm": 21.53518295288086,
+      "learning_rate": 3.2022022022022026e-06,
+      "loss": 0.3477,
+      "step": 67910
+    },
+    {
+      "epoch": 203.96,
+      "grad_norm": 17.988300323486328,
+      "learning_rate": 3.2012012012012016e-06,
+      "loss": 0.3504,
+      "step": 67920
+    },
+    {
+      "epoch": 203.99,
+      "grad_norm": 16.3185977935791,
+      "learning_rate": 3.2002002002002002e-06,
+      "loss": 0.3685,
+      "step": 67930
+    },
+    {
+      "epoch": 204.0,
+      "eval_accuracy": 0.8629,
+      "eval_loss": 0.5405135750770569,
+      "eval_runtime": 12.7245,
+      "eval_samples_per_second": 785.885,
+      "eval_steps_per_second": 3.144,
+      "step": 67932
+    },
+    {
+      "epoch": 204.02,
+      "grad_norm": 15.392194747924805,
+      "learning_rate": 3.1991991991991992e-06,
+      "loss": 0.3155,
+      "step": 67940
+    },
+    {
+      "epoch": 204.05,
+      "grad_norm": 14.565339088439941,
+      "learning_rate": 3.1981981981981987e-06,
+      "loss": 0.3779,
+      "step": 67950
+    },
+    {
+      "epoch": 204.08,
+      "grad_norm": 16.770187377929688,
+      "learning_rate": 3.1971971971971977e-06,
+      "loss": 0.3593,
+      "step": 67960
+    },
+    {
+      "epoch": 204.11,
+      "grad_norm": 16.89728546142578,
+      "learning_rate": 3.1961961961961963e-06,
+      "loss": 0.3773,
+      "step": 67970
+    },
+    {
+      "epoch": 204.14,
+      "grad_norm": 13.090940475463867,
+      "learning_rate": 3.1951951951951953e-06,
+      "loss": 0.3167,
+      "step": 67980
+    },
+    {
+      "epoch": 204.17,
+      "grad_norm": 21.46582794189453,
+      "learning_rate": 3.1941941941941943e-06,
+      "loss": 0.3986,
+      "step": 67990
+    },
+    {
+      "epoch": 204.2,
+      "grad_norm": 16.317481994628906,
+      "learning_rate": 3.1931931931931938e-06,
+      "loss": 0.3843,
+      "step": 68000
+    },
+    {
+      "epoch": 204.23,
+      "grad_norm": 21.513355255126953,
+      "learning_rate": 3.1921921921921924e-06,
+      "loss": 0.3416,
+      "step": 68010
+    },
+    {
+      "epoch": 204.26,
+      "grad_norm": 15.76412296295166,
+      "learning_rate": 3.1911911911911914e-06,
+      "loss": 0.3529,
+      "step": 68020
+    },
+    {
+      "epoch": 204.29,
+      "grad_norm": 16.10228729248047,
+      "learning_rate": 3.1901901901901904e-06,
+      "loss": 0.3284,
+      "step": 68030
+    },
+    {
+      "epoch": 204.32,
+      "grad_norm": 20.93988609313965,
+      "learning_rate": 3.1891891891891894e-06,
+      "loss": 0.3448,
+      "step": 68040
+    },
+    {
+      "epoch": 204.35,
+      "grad_norm": 11.768725395202637,
+      "learning_rate": 3.188188188188188e-06,
+      "loss": 0.3544,
+      "step": 68050
+    },
+    {
+      "epoch": 204.38,
+      "grad_norm": 18.16387367248535,
+      "learning_rate": 3.1871871871871875e-06,
+      "loss": 0.3376,
+      "step": 68060
+    },
+    {
+      "epoch": 204.41,
+      "grad_norm": 19.774927139282227,
+      "learning_rate": 3.1861861861861865e-06,
+      "loss": 0.3562,
+      "step": 68070
+    },
+    {
+      "epoch": 204.44,
+      "grad_norm": 22.80330467224121,
+      "learning_rate": 3.1851851851851855e-06,
+      "loss": 0.3503,
+      "step": 68080
+    },
+    {
+      "epoch": 204.47,
+      "grad_norm": 13.114124298095703,
+      "learning_rate": 3.184184184184184e-06,
+      "loss": 0.3328,
+      "step": 68090
+    },
+    {
+      "epoch": 204.5,
+      "grad_norm": 14.968913078308105,
+      "learning_rate": 3.183183183183183e-06,
+      "loss": 0.2999,
+      "step": 68100
+    },
+    {
+      "epoch": 204.53,
+      "grad_norm": 16.321569442749023,
+      "learning_rate": 3.1821821821821826e-06,
+      "loss": 0.3406,
+      "step": 68110
+    },
+    {
+      "epoch": 204.56,
+      "grad_norm": 16.081069946289062,
+      "learning_rate": 3.1811811811811816e-06,
+      "loss": 0.323,
+      "step": 68120
+    },
+    {
+      "epoch": 204.59,
+      "grad_norm": 25.609922409057617,
+      "learning_rate": 3.1801801801801806e-06,
+      "loss": 0.3744,
+      "step": 68130
+    },
+    {
+      "epoch": 204.62,
+      "grad_norm": 18.43142318725586,
+      "learning_rate": 3.179179179179179e-06,
+      "loss": 0.3658,
+      "step": 68140
+    },
+    {
+      "epoch": 204.65,
+      "grad_norm": 12.094287872314453,
+      "learning_rate": 3.1781781781781786e-06,
+      "loss": 0.3755,
+      "step": 68150
+    },
+    {
+      "epoch": 204.68,
+      "grad_norm": 12.684632301330566,
+      "learning_rate": 3.1771771771771776e-06,
+      "loss": 0.3376,
+      "step": 68160
+    },
+    {
+      "epoch": 204.71,
+      "grad_norm": 14.497475624084473,
+      "learning_rate": 3.1761761761761767e-06,
+      "loss": 0.3606,
+      "step": 68170
+    },
+    {
+      "epoch": 204.74,
+      "grad_norm": 10.236006736755371,
+      "learning_rate": 3.1751751751751753e-06,
+      "loss": 0.3046,
+      "step": 68180
+    },
+    {
+      "epoch": 204.77,
+      "grad_norm": 16.87126922607422,
+      "learning_rate": 3.1741741741741743e-06,
+      "loss": 0.3529,
+      "step": 68190
+    },
+    {
+      "epoch": 204.8,
+      "grad_norm": 15.732086181640625,
+      "learning_rate": 3.1731731731731737e-06,
+      "loss": 0.3238,
+      "step": 68200
+    },
+    {
+      "epoch": 204.83,
+      "grad_norm": 21.094053268432617,
+      "learning_rate": 3.1721721721721727e-06,
+      "loss": 0.3785,
+      "step": 68210
+    },
+    {
+      "epoch": 204.86,
+      "grad_norm": 17.592361450195312,
+      "learning_rate": 3.1711711711711713e-06,
+      "loss": 0.3735,
+      "step": 68220
+    },
+    {
+      "epoch": 204.89,
+      "grad_norm": 12.281081199645996,
+      "learning_rate": 3.1701701701701703e-06,
+      "loss": 0.4177,
+      "step": 68230
+    },
+    {
+      "epoch": 204.92,
+      "grad_norm": 13.502802848815918,
+      "learning_rate": 3.1691691691691694e-06,
+      "loss": 0.3053,
+      "step": 68240
+    },
+    {
+      "epoch": 204.95,
+      "grad_norm": 18.49332046508789,
+      "learning_rate": 3.168168168168169e-06,
+      "loss": 0.3012,
+      "step": 68250
+    },
+    {
+      "epoch": 204.98,
+      "grad_norm": 12.538700103759766,
+      "learning_rate": 3.1671671671671674e-06,
+      "loss": 0.3252,
+      "step": 68260
+    },
+    {
+      "epoch": 205.0,
+      "eval_accuracy": 0.8628,
+      "eval_loss": 0.541103720664978,
+      "eval_runtime": 12.9251,
+      "eval_samples_per_second": 773.691,
+      "eval_steps_per_second": 3.095,
+      "step": 68265
+    },
+    {
+      "epoch": 205.02,
+      "grad_norm": 12.407696723937988,
+      "learning_rate": 3.1661661661661664e-06,
+      "loss": 0.4103,
+      "step": 68270
+    },
+    {
+      "epoch": 205.05,
+      "grad_norm": 11.735945701599121,
+      "learning_rate": 3.1651651651651654e-06,
+      "loss": 0.3259,
+      "step": 68280
+    },
+    {
+      "epoch": 205.08,
+      "grad_norm": 14.637275695800781,
+      "learning_rate": 3.1641641641641645e-06,
+      "loss": 0.3928,
+      "step": 68290
+    },
+    {
+      "epoch": 205.11,
+      "grad_norm": 16.591323852539062,
+      "learning_rate": 3.163163163163163e-06,
+      "loss": 0.349,
+      "step": 68300
+    },
+    {
+      "epoch": 205.14,
+      "grad_norm": 21.172943115234375,
+      "learning_rate": 3.1621621621621625e-06,
+      "loss": 0.3677,
+      "step": 68310
+    },
+    {
+      "epoch": 205.17,
+      "grad_norm": 15.315435409545898,
+      "learning_rate": 3.1611611611611615e-06,
+      "loss": 0.3446,
+      "step": 68320
+    },
+    {
+      "epoch": 205.2,
+      "grad_norm": 18.649477005004883,
+      "learning_rate": 3.1601601601601605e-06,
+      "loss": 0.3174,
+      "step": 68330
+    },
+    {
+      "epoch": 205.23,
+      "grad_norm": 15.136581420898438,
+      "learning_rate": 3.159159159159159e-06,
+      "loss": 0.366,
+      "step": 68340
+    },
+    {
+      "epoch": 205.26,
+      "grad_norm": 13.142144203186035,
+      "learning_rate": 3.158158158158158e-06,
+      "loss": 0.4016,
+      "step": 68350
+    },
+    {
+      "epoch": 205.29,
+      "grad_norm": 18.103031158447266,
+      "learning_rate": 3.1571571571571576e-06,
+      "loss": 0.3613,
+      "step": 68360
+    },
+    {
+      "epoch": 205.32,
+      "grad_norm": 14.686224937438965,
+      "learning_rate": 3.1561561561561566e-06,
+      "loss": 0.3752,
+      "step": 68370
+    },
+    {
+      "epoch": 205.35,
+      "grad_norm": 17.402305603027344,
+      "learning_rate": 3.155155155155155e-06,
+      "loss": 0.3215,
+      "step": 68380
+    },
+    {
+      "epoch": 205.38,
+      "grad_norm": 20.296112060546875,
+      "learning_rate": 3.1541541541541542e-06,
+      "loss": 0.3846,
+      "step": 68390
+    },
+    {
+      "epoch": 205.41,
+      "grad_norm": 14.213618278503418,
+      "learning_rate": 3.1531531531531532e-06,
+      "loss": 0.3679,
+      "step": 68400
+    },
+    {
+      "epoch": 205.44,
+      "grad_norm": 19.24360466003418,
+      "learning_rate": 3.1521521521521527e-06,
+      "loss": 0.3505,
+      "step": 68410
+    },
+    {
+      "epoch": 205.47,
+      "grad_norm": 24.330955505371094,
+      "learning_rate": 3.1511511511511517e-06,
+      "loss": 0.337,
+      "step": 68420
+    },
+    {
+      "epoch": 205.5,
+      "grad_norm": 18.127689361572266,
+      "learning_rate": 3.1501501501501503e-06,
+      "loss": 0.3281,
+      "step": 68430
+    },
+    {
+      "epoch": 205.53,
+      "grad_norm": 10.580160140991211,
+      "learning_rate": 3.1491491491491493e-06,
+      "loss": 0.3421,
+      "step": 68440
+    },
+    {
+      "epoch": 205.56,
+      "grad_norm": 23.078493118286133,
+      "learning_rate": 3.1481481481481483e-06,
+      "loss": 0.338,
+      "step": 68450
+    },
+    {
+      "epoch": 205.59,
+      "grad_norm": 22.84177017211914,
+      "learning_rate": 3.1471471471471478e-06,
+      "loss": 0.363,
+      "step": 68460
+    },
+    {
+      "epoch": 205.62,
+      "grad_norm": 18.662574768066406,
+      "learning_rate": 3.1461461461461464e-06,
+      "loss": 0.3549,
+      "step": 68470
+    },
+    {
+      "epoch": 205.65,
+      "grad_norm": 12.01109790802002,
+      "learning_rate": 3.1451451451451454e-06,
+      "loss": 0.3481,
+      "step": 68480
+    },
+    {
+      "epoch": 205.68,
+      "grad_norm": 17.017311096191406,
+      "learning_rate": 3.1441441441441444e-06,
+      "loss": 0.3275,
+      "step": 68490
+    },
+    {
+      "epoch": 205.71,
+      "grad_norm": 23.50739288330078,
+      "learning_rate": 3.1431431431431434e-06,
+      "loss": 0.368,
+      "step": 68500
+    },
+    {
+      "epoch": 205.74,
+      "grad_norm": 20.995338439941406,
+      "learning_rate": 3.142142142142142e-06,
+      "loss": 0.3235,
+      "step": 68510
+    },
+    {
+      "epoch": 205.77,
+      "grad_norm": 15.138721466064453,
+      "learning_rate": 3.1411411411411415e-06,
+      "loss": 0.3513,
+      "step": 68520
+    },
+    {
+      "epoch": 205.8,
+      "grad_norm": 19.11183738708496,
+      "learning_rate": 3.1401401401401405e-06,
+      "loss": 0.3863,
+      "step": 68530
+    },
+    {
+      "epoch": 205.83,
+      "grad_norm": 25.361576080322266,
+      "learning_rate": 3.1391391391391395e-06,
+      "loss": 0.3427,
+      "step": 68540
+    },
+    {
+      "epoch": 205.86,
+      "grad_norm": 13.447063446044922,
+      "learning_rate": 3.138138138138138e-06,
+      "loss": 0.3762,
+      "step": 68550
+    },
+    {
+      "epoch": 205.89,
+      "grad_norm": 10.898679733276367,
+      "learning_rate": 3.137137137137137e-06,
+      "loss": 0.3692,
+      "step": 68560
+    },
+    {
+      "epoch": 205.92,
+      "grad_norm": 13.228545188903809,
+      "learning_rate": 3.1361361361361365e-06,
+      "loss": 0.3105,
+      "step": 68570
+    },
+    {
+      "epoch": 205.95,
+      "grad_norm": 16.638708114624023,
+      "learning_rate": 3.1351351351351356e-06,
+      "loss": 0.394,
+      "step": 68580
+    },
+    {
+      "epoch": 205.98,
+      "grad_norm": 12.935309410095215,
+      "learning_rate": 3.134134134134134e-06,
+      "loss": 0.358,
+      "step": 68590
+    },
+    {
+      "epoch": 206.0,
+      "eval_accuracy": 0.8621,
+      "eval_loss": 0.5403095483779907,
+      "eval_runtime": 12.6015,
+      "eval_samples_per_second": 793.554,
+      "eval_steps_per_second": 3.174,
+      "step": 68598
+    },
+    {
+      "epoch": 206.01,
+      "grad_norm": 18.407442092895508,
+      "learning_rate": 3.133133133133133e-06,
+      "loss": 0.503,
+      "step": 68600
+    },
+    {
+      "epoch": 206.04,
+      "grad_norm": 18.42306900024414,
+      "learning_rate": 3.132132132132132e-06,
+      "loss": 0.3331,
+      "step": 68610
+    },
+    {
+      "epoch": 206.07,
+      "grad_norm": 11.975234031677246,
+      "learning_rate": 3.1311311311311316e-06,
+      "loss": 0.3641,
+      "step": 68620
+    },
+    {
+      "epoch": 206.1,
+      "grad_norm": 11.405656814575195,
+      "learning_rate": 3.1301301301301302e-06,
+      "loss": 0.3552,
+      "step": 68630
+    },
+    {
+      "epoch": 206.13,
+      "grad_norm": 15.915101051330566,
+      "learning_rate": 3.1291291291291293e-06,
+      "loss": 0.3347,
+      "step": 68640
+    },
+    {
+      "epoch": 206.16,
+      "grad_norm": 16.570449829101562,
+      "learning_rate": 3.1281281281281283e-06,
+      "loss": 0.3288,
+      "step": 68650
+    },
+    {
+      "epoch": 206.19,
+      "grad_norm": 16.335290908813477,
+      "learning_rate": 3.1271271271271277e-06,
+      "loss": 0.372,
+      "step": 68660
+    },
+    {
+      "epoch": 206.22,
+      "grad_norm": 9.110803604125977,
+      "learning_rate": 3.1261261261261267e-06,
+      "loss": 0.3347,
+      "step": 68670
+    },
+    {
+      "epoch": 206.25,
+      "grad_norm": 17.05145835876465,
+      "learning_rate": 3.1251251251251253e-06,
+      "loss": 0.3266,
+      "step": 68680
+    },
+    {
+      "epoch": 206.28,
+      "grad_norm": 22.380210876464844,
+      "learning_rate": 3.1241241241241243e-06,
+      "loss": 0.3465,
+      "step": 68690
+    },
+    {
+      "epoch": 206.31,
+      "grad_norm": 12.037788391113281,
+      "learning_rate": 3.1231231231231234e-06,
+      "loss": 0.3508,
+      "step": 68700
+    },
+    {
+      "epoch": 206.34,
+      "grad_norm": 24.467702865600586,
+      "learning_rate": 3.122122122122123e-06,
+      "loss": 0.3549,
+      "step": 68710
+    },
+    {
+      "epoch": 206.37,
+      "grad_norm": 15.965084075927734,
+      "learning_rate": 3.1211211211211214e-06,
+      "loss": 0.2974,
+      "step": 68720
+    },
+    {
+      "epoch": 206.4,
+      "grad_norm": 18.54871368408203,
+      "learning_rate": 3.1201201201201204e-06,
+      "loss": 0.3709,
+      "step": 68730
+    },
+    {
+      "epoch": 206.43,
+      "grad_norm": 16.23748207092285,
+      "learning_rate": 3.1191191191191194e-06,
+      "loss": 0.3877,
+      "step": 68740
+    },
+    {
+      "epoch": 206.46,
+      "grad_norm": 15.845029830932617,
+      "learning_rate": 3.1181181181181185e-06,
+      "loss": 0.3702,
+      "step": 68750
+    },
+    {
+      "epoch": 206.49,
+      "grad_norm": 15.77956771850586,
+      "learning_rate": 3.117117117117117e-06,
+      "loss": 0.3793,
+      "step": 68760
+    },
+    {
+      "epoch": 206.52,
+      "grad_norm": 12.40796184539795,
+      "learning_rate": 3.1161161161161165e-06,
+      "loss": 0.3669,
+      "step": 68770
+    },
+    {
+      "epoch": 206.55,
+      "grad_norm": 12.436490058898926,
+      "learning_rate": 3.1151151151151155e-06,
+      "loss": 0.3546,
+      "step": 68780
+    },
+    {
+      "epoch": 206.58,
+      "grad_norm": 19.976926803588867,
+      "learning_rate": 3.1141141141141145e-06,
+      "loss": 0.3441,
+      "step": 68790
+    },
+    {
+      "epoch": 206.61,
+      "grad_norm": 18.680728912353516,
+      "learning_rate": 3.113113113113113e-06,
+      "loss": 0.3261,
+      "step": 68800
+    },
+    {
+      "epoch": 206.64,
+      "grad_norm": 14.421867370605469,
+      "learning_rate": 3.112112112112112e-06,
+      "loss": 0.365,
+      "step": 68810
+    },
+    {
+      "epoch": 206.67,
+      "grad_norm": 12.137272834777832,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 0.3534,
+      "step": 68820
+    },
+    {
+      "epoch": 206.7,
+      "grad_norm": 20.421470642089844,
+      "learning_rate": 3.1101101101101106e-06,
+      "loss": 0.3558,
+      "step": 68830
+    },
+    {
+      "epoch": 206.73,
+      "grad_norm": 17.597305297851562,
+      "learning_rate": 3.109109109109109e-06,
+      "loss": 0.4026,
+      "step": 68840
+    },
+    {
+      "epoch": 206.76,
+      "grad_norm": 28.643882751464844,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.3997,
+      "step": 68850
+    },
+    {
+      "epoch": 206.79,
+      "grad_norm": 15.382844924926758,
+      "learning_rate": 3.1071071071071072e-06,
+      "loss": 0.3485,
+      "step": 68860
+    },
+    {
+      "epoch": 206.82,
+      "grad_norm": 15.038064956665039,
+      "learning_rate": 3.1061061061061067e-06,
+      "loss": 0.3379,
+      "step": 68870
+    },
+    {
+      "epoch": 206.85,
+      "grad_norm": 14.94677448272705,
+      "learning_rate": 3.1051051051051053e-06,
+      "loss": 0.3336,
+      "step": 68880
+    },
+    {
+      "epoch": 206.88,
+      "grad_norm": 14.845500946044922,
+      "learning_rate": 3.1041041041041043e-06,
+      "loss": 0.3885,
+      "step": 68890
+    },
+    {
+      "epoch": 206.91,
+      "grad_norm": 15.164229393005371,
+      "learning_rate": 3.1031031031031033e-06,
+      "loss": 0.3487,
+      "step": 68900
+    },
+    {
+      "epoch": 206.94,
+      "grad_norm": 18.9834041595459,
+      "learning_rate": 3.1021021021021023e-06,
+      "loss": 0.3309,
+      "step": 68910
+    },
+    {
+      "epoch": 206.97,
+      "grad_norm": 15.080185890197754,
+      "learning_rate": 3.1011011011011018e-06,
+      "loss": 0.3744,
+      "step": 68920
+    },
+    {
+      "epoch": 207.0,
+      "grad_norm": 17.9138126373291,
+      "learning_rate": 3.1001001001001004e-06,
+      "loss": 0.3086,
+      "step": 68930
+    },
+    {
+      "epoch": 207.0,
+      "eval_accuracy": 0.8626,
+      "eval_loss": 0.5398733615875244,
+      "eval_runtime": 12.7427,
+      "eval_samples_per_second": 784.76,
+      "eval_steps_per_second": 3.139,
+      "step": 68931
+    },
+    {
+      "epoch": 207.03,
+      "grad_norm": 20.874412536621094,
+      "learning_rate": 3.0990990990990994e-06,
+      "loss": 0.3598,
+      "step": 68940
+    },
+    {
+      "epoch": 207.06,
+      "grad_norm": 17.016088485717773,
+      "learning_rate": 3.0980980980980984e-06,
+      "loss": 0.3672,
+      "step": 68950
+    },
+    {
+      "epoch": 207.09,
+      "grad_norm": 16.138824462890625,
+      "learning_rate": 3.0970970970970974e-06,
+      "loss": 0.3595,
+      "step": 68960
+    },
+    {
+      "epoch": 207.12,
+      "grad_norm": 13.253243446350098,
+      "learning_rate": 3.096096096096096e-06,
+      "loss": 0.3669,
+      "step": 68970
+    },
+    {
+      "epoch": 207.15,
+      "grad_norm": 19.4034481048584,
+      "learning_rate": 3.0950950950950955e-06,
+      "loss": 0.3939,
+      "step": 68980
+    },
+    {
+      "epoch": 207.18,
+      "grad_norm": 9.029620170593262,
+      "learning_rate": 3.0940940940940945e-06,
+      "loss": 0.323,
+      "step": 68990
+    },
+    {
+      "epoch": 207.21,
+      "grad_norm": 13.628242492675781,
+      "learning_rate": 3.0930930930930935e-06,
+      "loss": 0.3741,
+      "step": 69000
+    },
+    {
+      "epoch": 207.24,
+      "grad_norm": 13.685140609741211,
+      "learning_rate": 3.092092092092092e-06,
+      "loss": 0.3282,
+      "step": 69010
+    },
+    {
+      "epoch": 207.27,
+      "grad_norm": 23.7899227142334,
+      "learning_rate": 3.091091091091091e-06,
+      "loss": 0.3906,
+      "step": 69020
+    },
+    {
+      "epoch": 207.3,
+      "grad_norm": 26.971893310546875,
+      "learning_rate": 3.0900900900900905e-06,
+      "loss": 0.382,
+      "step": 69030
+    },
+    {
+      "epoch": 207.33,
+      "grad_norm": 15.214637756347656,
+      "learning_rate": 3.0890890890890896e-06,
+      "loss": 0.3254,
+      "step": 69040
+    },
+    {
+      "epoch": 207.36,
+      "grad_norm": 18.76878547668457,
+      "learning_rate": 3.088088088088088e-06,
+      "loss": 0.3641,
+      "step": 69050
+    },
+    {
+      "epoch": 207.39,
+      "grad_norm": 9.363194465637207,
+      "learning_rate": 3.087087087087087e-06,
+      "loss": 0.3271,
+      "step": 69060
+    },
+    {
+      "epoch": 207.42,
+      "grad_norm": 10.995755195617676,
+      "learning_rate": 3.086086086086086e-06,
+      "loss": 0.3366,
+      "step": 69070
+    },
+    {
+      "epoch": 207.45,
+      "grad_norm": 16.75929832458496,
+      "learning_rate": 3.0850850850850856e-06,
+      "loss": 0.3643,
+      "step": 69080
+    },
+    {
+      "epoch": 207.48,
+      "grad_norm": 12.146560668945312,
+      "learning_rate": 3.0840840840840842e-06,
+      "loss": 0.3244,
+      "step": 69090
+    },
+    {
+      "epoch": 207.51,
+      "grad_norm": 14.127967834472656,
+      "learning_rate": 3.0830830830830832e-06,
+      "loss": 0.3726,
+      "step": 69100
+    },
+    {
+      "epoch": 207.54,
+      "grad_norm": 14.729283332824707,
+      "learning_rate": 3.0820820820820823e-06,
+      "loss": 0.3489,
+      "step": 69110
+    },
+    {
+      "epoch": 207.57,
+      "grad_norm": 21.01169204711914,
+      "learning_rate": 3.0810810810810817e-06,
+      "loss": 0.361,
+      "step": 69120
+    },
+    {
+      "epoch": 207.6,
+      "grad_norm": 15.380688667297363,
+      "learning_rate": 3.08008008008008e-06,
+      "loss": 0.319,
+      "step": 69130
+    },
+    {
+      "epoch": 207.63,
+      "grad_norm": 13.213290214538574,
+      "learning_rate": 3.0790790790790793e-06,
+      "loss": 0.3287,
+      "step": 69140
+    },
+    {
+      "epoch": 207.66,
+      "grad_norm": 9.306558609008789,
+      "learning_rate": 3.0780780780780783e-06,
+      "loss": 0.4083,
+      "step": 69150
+    },
+    {
+      "epoch": 207.69,
+      "grad_norm": 19.454538345336914,
+      "learning_rate": 3.0770770770770774e-06,
+      "loss": 0.4142,
+      "step": 69160
+    },
+    {
+      "epoch": 207.72,
+      "grad_norm": 14.913165092468262,
+      "learning_rate": 3.076076076076076e-06,
+      "loss": 0.3498,
+      "step": 69170
+    },
+    {
+      "epoch": 207.75,
+      "grad_norm": 19.51289939880371,
+      "learning_rate": 3.0750750750750754e-06,
+      "loss": 0.357,
+      "step": 69180
+    },
+    {
+      "epoch": 207.78,
+      "grad_norm": 16.138532638549805,
+      "learning_rate": 3.0740740740740744e-06,
+      "loss": 0.3568,
+      "step": 69190
+    },
+    {
+      "epoch": 207.81,
+      "grad_norm": 20.34121322631836,
+      "learning_rate": 3.0730730730730734e-06,
+      "loss": 0.3135,
+      "step": 69200
+    },
+    {
+      "epoch": 207.84,
+      "grad_norm": 11.682665824890137,
+      "learning_rate": 3.0720720720720725e-06,
+      "loss": 0.3341,
+      "step": 69210
+    },
+    {
+      "epoch": 207.87,
+      "grad_norm": 18.445932388305664,
+      "learning_rate": 3.071071071071071e-06,
+      "loss": 0.3707,
+      "step": 69220
+    },
+    {
+      "epoch": 207.9,
+      "grad_norm": 21.723405838012695,
+      "learning_rate": 3.0700700700700705e-06,
+      "loss": 0.3645,
+      "step": 69230
+    },
+    {
+      "epoch": 207.93,
+      "grad_norm": 12.624300956726074,
+      "learning_rate": 3.0690690690690695e-06,
+      "loss": 0.3238,
+      "step": 69240
+    },
+    {
+      "epoch": 207.96,
+      "grad_norm": 24.636821746826172,
+      "learning_rate": 3.0680680680680685e-06,
+      "loss": 0.3628,
+      "step": 69250
+    },
+    {
+      "epoch": 207.99,
+      "grad_norm": 21.74216651916504,
+      "learning_rate": 3.067067067067067e-06,
+      "loss": 0.3774,
+      "step": 69260
+    },
+    {
+      "epoch": 208.0,
+      "eval_accuracy": 0.8628,
+      "eval_loss": 0.5390221476554871,
+      "eval_runtime": 12.7397,
+      "eval_samples_per_second": 784.95,
+      "eval_steps_per_second": 3.14,
+      "step": 69264
+    },
+    {
+      "epoch": 208.02,
+      "grad_norm": 14.526623725891113,
+      "learning_rate": 3.066066066066066e-06,
+      "loss": 0.3168,
+      "step": 69270
+    },
+    {
+      "epoch": 208.05,
+      "grad_norm": 12.639586448669434,
+      "learning_rate": 3.0650650650650656e-06,
+      "loss": 0.3252,
+      "step": 69280
+    },
+    {
+      "epoch": 208.08,
+      "grad_norm": 11.387578010559082,
+      "learning_rate": 3.0640640640640646e-06,
+      "loss": 0.3477,
+      "step": 69290
+    },
+    {
+      "epoch": 208.11,
+      "grad_norm": 12.791800498962402,
+      "learning_rate": 3.063063063063063e-06,
+      "loss": 0.3103,
+      "step": 69300
+    },
+    {
+      "epoch": 208.14,
+      "grad_norm": 12.370397567749023,
+      "learning_rate": 3.062062062062062e-06,
+      "loss": 0.3541,
+      "step": 69310
+    },
+    {
+      "epoch": 208.17,
+      "grad_norm": 15.504064559936523,
+      "learning_rate": 3.0610610610610612e-06,
+      "loss": 0.3346,
+      "step": 69320
+    },
+    {
+      "epoch": 208.2,
+      "grad_norm": 16.819915771484375,
+      "learning_rate": 3.0600600600600607e-06,
+      "loss": 0.3361,
+      "step": 69330
+    },
+    {
+      "epoch": 208.23,
+      "grad_norm": 15.010241508483887,
+      "learning_rate": 3.0590590590590593e-06,
+      "loss": 0.3512,
+      "step": 69340
+    },
+    {
+      "epoch": 208.26,
+      "grad_norm": 18.151119232177734,
+      "learning_rate": 3.0580580580580583e-06,
+      "loss": 0.3579,
+      "step": 69350
+    },
+    {
+      "epoch": 208.29,
+      "grad_norm": 15.59250259399414,
+      "learning_rate": 3.0570570570570573e-06,
+      "loss": 0.3543,
+      "step": 69360
+    },
+    {
+      "epoch": 208.32,
+      "grad_norm": 12.143160820007324,
+      "learning_rate": 3.0560560560560563e-06,
+      "loss": 0.3587,
+      "step": 69370
+    },
+    {
+      "epoch": 208.35,
+      "grad_norm": 20.135801315307617,
+      "learning_rate": 3.055055055055055e-06,
+      "loss": 0.3605,
+      "step": 69380
+    },
+    {
+      "epoch": 208.38,
+      "grad_norm": 16.066307067871094,
+      "learning_rate": 3.0540540540540544e-06,
+      "loss": 0.3336,
+      "step": 69390
+    },
+    {
+      "epoch": 208.41,
+      "grad_norm": 11.094693183898926,
+      "learning_rate": 3.0530530530530534e-06,
+      "loss": 0.3282,
+      "step": 69400
+    },
+    {
+      "epoch": 208.44,
+      "grad_norm": 14.357072830200195,
+      "learning_rate": 3.0520520520520524e-06,
+      "loss": 0.3109,
+      "step": 69410
+    },
+    {
+      "epoch": 208.47,
+      "grad_norm": 15.544072151184082,
+      "learning_rate": 3.051051051051051e-06,
+      "loss": 0.3328,
+      "step": 69420
+    },
+    {
+      "epoch": 208.5,
+      "grad_norm": 14.240755081176758,
+      "learning_rate": 3.05005005005005e-06,
+      "loss": 0.3144,
+      "step": 69430
+    },
+    {
+      "epoch": 208.53,
+      "grad_norm": 16.881847381591797,
+      "learning_rate": 3.0490490490490494e-06,
+      "loss": 0.305,
+      "step": 69440
+    },
+    {
+      "epoch": 208.56,
+      "grad_norm": 13.087959289550781,
+      "learning_rate": 3.0480480480480485e-06,
+      "loss": 0.3552,
+      "step": 69450
+    },
+    {
+      "epoch": 208.59,
+      "grad_norm": 25.458900451660156,
+      "learning_rate": 3.0470470470470475e-06,
+      "loss": 0.3236,
+      "step": 69460
+    },
+    {
+      "epoch": 208.62,
+      "grad_norm": 24.415706634521484,
+      "learning_rate": 3.046046046046046e-06,
+      "loss": 0.3378,
+      "step": 69470
+    },
+    {
+      "epoch": 208.65,
+      "grad_norm": 15.777860641479492,
+      "learning_rate": 3.045045045045045e-06,
+      "loss": 0.3796,
+      "step": 69480
+    },
+    {
+      "epoch": 208.68,
+      "grad_norm": 21.747879028320312,
+      "learning_rate": 3.0440440440440445e-06,
+      "loss": 0.3293,
+      "step": 69490
+    },
+    {
+      "epoch": 208.71,
+      "grad_norm": 19.410884857177734,
+      "learning_rate": 3.0430430430430436e-06,
+      "loss": 0.3442,
+      "step": 69500
+    },
+    {
+      "epoch": 208.74,
+      "grad_norm": 34.77320861816406,
+      "learning_rate": 3.042042042042042e-06,
+      "loss": 0.3648,
+      "step": 69510
+    },
+    {
+      "epoch": 208.77,
+      "grad_norm": 11.096572875976562,
+      "learning_rate": 3.041041041041041e-06,
+      "loss": 0.356,
+      "step": 69520
+    },
+    {
+      "epoch": 208.8,
+      "grad_norm": 18.793962478637695,
+      "learning_rate": 3.04004004004004e-06,
+      "loss": 0.3578,
+      "step": 69530
+    },
+    {
+      "epoch": 208.83,
+      "grad_norm": 20.74724578857422,
+      "learning_rate": 3.0390390390390396e-06,
+      "loss": 0.3343,
+      "step": 69540
+    },
+    {
+      "epoch": 208.86,
+      "grad_norm": 14.952449798583984,
+      "learning_rate": 3.0380380380380382e-06,
+      "loss": 0.3191,
+      "step": 69550
+    },
+    {
+      "epoch": 208.89,
+      "grad_norm": 22.39580535888672,
+      "learning_rate": 3.0370370370370372e-06,
+      "loss": 0.3576,
+      "step": 69560
+    },
+    {
+      "epoch": 208.92,
+      "grad_norm": 14.421562194824219,
+      "learning_rate": 3.0360360360360363e-06,
+      "loss": 0.4019,
+      "step": 69570
+    },
+    {
+      "epoch": 208.95,
+      "grad_norm": 18.53263282775879,
+      "learning_rate": 3.0350350350350353e-06,
+      "loss": 0.3308,
+      "step": 69580
+    },
+    {
+      "epoch": 208.98,
+      "grad_norm": 14.5794038772583,
+      "learning_rate": 3.034034034034034e-06,
+      "loss": 0.3449,
+      "step": 69590
+    },
+    {
+      "epoch": 209.0,
+      "eval_accuracy": 0.865,
+      "eval_loss": 0.5388179421424866,
+      "eval_runtime": 12.84,
+      "eval_samples_per_second": 778.818,
+      "eval_steps_per_second": 3.115,
+      "step": 69597
+    },
+    {
+      "epoch": 209.01,
+      "grad_norm": 17.315523147583008,
+      "learning_rate": 3.0330330330330333e-06,
+      "loss": 0.3138,
+      "step": 69600
+    },
+    {
+      "epoch": 209.04,
+      "grad_norm": 11.96428394317627,
+      "learning_rate": 3.0320320320320323e-06,
+      "loss": 0.3073,
+      "step": 69610
+    },
+    {
+      "epoch": 209.07,
+      "grad_norm": 20.06438636779785,
+      "learning_rate": 3.0310310310310314e-06,
+      "loss": 0.358,
+      "step": 69620
+    },
+    {
+      "epoch": 209.1,
+      "grad_norm": 10.198114395141602,
+      "learning_rate": 3.03003003003003e-06,
+      "loss": 0.313,
+      "step": 69630
+    },
+    {
+      "epoch": 209.13,
+      "grad_norm": 11.16504192352295,
+      "learning_rate": 3.029029029029029e-06,
+      "loss": 0.3249,
+      "step": 69640
+    },
+    {
+      "epoch": 209.16,
+      "grad_norm": 15.243680000305176,
+      "learning_rate": 3.0280280280280284e-06,
+      "loss": 0.3045,
+      "step": 69650
+    },
+    {
+      "epoch": 209.19,
+      "grad_norm": 19.457910537719727,
+      "learning_rate": 3.0270270270270274e-06,
+      "loss": 0.4001,
+      "step": 69660
+    },
+    {
+      "epoch": 209.22,
+      "grad_norm": 14.875446319580078,
+      "learning_rate": 3.026026026026026e-06,
+      "loss": 0.3557,
+      "step": 69670
+    },
+    {
+      "epoch": 209.25,
+      "grad_norm": 25.521038055419922,
+      "learning_rate": 3.025025025025025e-06,
+      "loss": 0.3316,
+      "step": 69680
+    },
+    {
+      "epoch": 209.28,
+      "grad_norm": 10.678447723388672,
+      "learning_rate": 3.0240240240240245e-06,
+      "loss": 0.3521,
+      "step": 69690
+    },
+    {
+      "epoch": 209.31,
+      "grad_norm": 17.137130737304688,
+      "learning_rate": 3.0230230230230235e-06,
+      "loss": 0.3617,
+      "step": 69700
+    },
+    {
+      "epoch": 209.34,
+      "grad_norm": 21.64864158630371,
+      "learning_rate": 3.022022022022022e-06,
+      "loss": 0.3346,
+      "step": 69710
+    },
+    {
+      "epoch": 209.37,
+      "grad_norm": 11.620427131652832,
+      "learning_rate": 3.021021021021021e-06,
+      "loss": 0.354,
+      "step": 69720
+    },
+    {
+      "epoch": 209.4,
+      "grad_norm": 13.872488021850586,
+      "learning_rate": 3.02002002002002e-06,
+      "loss": 0.3533,
+      "step": 69730
+    },
+    {
+      "epoch": 209.43,
+      "grad_norm": 16.69676971435547,
+      "learning_rate": 3.0190190190190196e-06,
+      "loss": 0.3676,
+      "step": 69740
+    },
+    {
+      "epoch": 209.46,
+      "grad_norm": 23.858285903930664,
+      "learning_rate": 3.0180180180180186e-06,
+      "loss": 0.3713,
+      "step": 69750
+    },
+    {
+      "epoch": 209.49,
+      "grad_norm": 22.945720672607422,
+      "learning_rate": 3.017017017017017e-06,
+      "loss": 0.3725,
+      "step": 69760
+    },
+    {
+      "epoch": 209.52,
+      "grad_norm": 16.24211311340332,
+      "learning_rate": 3.016016016016016e-06,
+      "loss": 0.3496,
+      "step": 69770
+    },
+    {
+      "epoch": 209.55,
+      "grad_norm": 16.850528717041016,
+      "learning_rate": 3.0150150150150152e-06,
+      "loss": 0.3503,
+      "step": 69780
+    },
+    {
+      "epoch": 209.58,
+      "grad_norm": 24.07161521911621,
+      "learning_rate": 3.0140140140140147e-06,
+      "loss": 0.3548,
+      "step": 69790
+    },
+    {
+      "epoch": 209.61,
+      "grad_norm": 17.17265510559082,
+      "learning_rate": 3.0130130130130133e-06,
+      "loss": 0.3077,
+      "step": 69800
+    },
+    {
+      "epoch": 209.64,
+      "grad_norm": 13.473716735839844,
+      "learning_rate": 3.0120120120120123e-06,
+      "loss": 0.3096,
+      "step": 69810
+    },
+    {
+      "epoch": 209.67,
+      "grad_norm": 17.760986328125,
+      "learning_rate": 3.0110110110110113e-06,
+      "loss": 0.3537,
+      "step": 69820
+    },
+    {
+      "epoch": 209.7,
+      "grad_norm": 18.669309616088867,
+      "learning_rate": 3.0100100100100103e-06,
+      "loss": 0.3199,
+      "step": 69830
+    },
+    {
+      "epoch": 209.73,
+      "grad_norm": 14.330032348632812,
+      "learning_rate": 3.009009009009009e-06,
+      "loss": 0.3527,
+      "step": 69840
+    },
+    {
+      "epoch": 209.76,
+      "grad_norm": 19.19539451599121,
+      "learning_rate": 3.0080080080080084e-06,
+      "loss": 0.3478,
+      "step": 69850
+    },
+    {
+      "epoch": 209.79,
+      "grad_norm": 15.003211975097656,
+      "learning_rate": 3.0070070070070074e-06,
+      "loss": 0.3187,
+      "step": 69860
+    },
+    {
+      "epoch": 209.82,
+      "grad_norm": 13.076828002929688,
+      "learning_rate": 3.0060060060060064e-06,
+      "loss": 0.3373,
+      "step": 69870
+    },
+    {
+      "epoch": 209.85,
+      "grad_norm": 14.027448654174805,
+      "learning_rate": 3.005005005005005e-06,
+      "loss": 0.3492,
+      "step": 69880
+    },
+    {
+      "epoch": 209.88,
+      "grad_norm": 16.81917953491211,
+      "learning_rate": 3.004004004004004e-06,
+      "loss": 0.3344,
+      "step": 69890
+    },
+    {
+      "epoch": 209.91,
+      "grad_norm": 26.22533416748047,
+      "learning_rate": 3.0030030030030034e-06,
+      "loss": 0.3351,
+      "step": 69900
+    },
+    {
+      "epoch": 209.94,
+      "grad_norm": 13.289993286132812,
+      "learning_rate": 3.0020020020020025e-06,
+      "loss": 0.3651,
+      "step": 69910
+    },
+    {
+      "epoch": 209.97,
+      "grad_norm": 18.05742073059082,
+      "learning_rate": 3.001001001001001e-06,
+      "loss": 0.3595,
+      "step": 69920
+    },
+    {
+      "epoch": 210.0,
+      "grad_norm": 2.621110677719116,
+      "learning_rate": 3e-06,
+      "loss": 0.3268,
+      "step": 69930
+    },
+    {
+      "epoch": 210.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5362868309020996,
+      "eval_runtime": 12.8628,
+      "eval_samples_per_second": 777.438,
+      "eval_steps_per_second": 3.11,
+      "step": 69930
+    },
+    {
+      "epoch": 210.03,
+      "grad_norm": 18.765581130981445,
+      "learning_rate": 2.998998998998999e-06,
+      "loss": 0.3644,
+      "step": 69940
+    },
+    {
+      "epoch": 210.06,
+      "grad_norm": 20.218130111694336,
+      "learning_rate": 2.9979979979979985e-06,
+      "loss": 0.3705,
+      "step": 69950
+    },
+    {
+      "epoch": 210.09,
+      "grad_norm": 11.384099960327148,
+      "learning_rate": 2.996996996996997e-06,
+      "loss": 0.3063,
+      "step": 69960
+    },
+    {
+      "epoch": 210.12,
+      "grad_norm": 23.85811424255371,
+      "learning_rate": 2.995995995995996e-06,
+      "loss": 0.3739,
+      "step": 69970
+    },
+    {
+      "epoch": 210.15,
+      "grad_norm": 29.71883201599121,
+      "learning_rate": 2.994994994994995e-06,
+      "loss": 0.3837,
+      "step": 69980
+    },
+    {
+      "epoch": 210.18,
+      "grad_norm": 13.400443077087402,
+      "learning_rate": 2.993993993993994e-06,
+      "loss": 0.3664,
+      "step": 69990
+    },
+    {
+      "epoch": 210.21,
+      "grad_norm": 13.40132999420166,
+      "learning_rate": 2.9929929929929936e-06,
+      "loss": 0.3309,
+      "step": 70000
+    },
+    {
+      "epoch": 210.24,
+      "grad_norm": 10.763825416564941,
+      "learning_rate": 2.9919919919919922e-06,
+      "loss": 0.3262,
+      "step": 70010
+    },
+    {
+      "epoch": 210.27,
+      "grad_norm": 18.24542236328125,
+      "learning_rate": 2.9909909909909912e-06,
+      "loss": 0.3005,
+      "step": 70020
+    },
+    {
+      "epoch": 210.3,
+      "grad_norm": 15.229015350341797,
+      "learning_rate": 2.9899899899899903e-06,
+      "loss": 0.3238,
+      "step": 70030
+    },
+    {
+      "epoch": 210.33,
+      "grad_norm": 13.976673126220703,
+      "learning_rate": 2.9889889889889893e-06,
+      "loss": 0.3949,
+      "step": 70040
+    },
+    {
+      "epoch": 210.36,
+      "grad_norm": 25.43150520324707,
+      "learning_rate": 2.987987987987988e-06,
+      "loss": 0.3772,
+      "step": 70050
+    },
+    {
+      "epoch": 210.39,
+      "grad_norm": 16.01929473876953,
+      "learning_rate": 2.9869869869869873e-06,
+      "loss": 0.3581,
+      "step": 70060
+    },
+    {
+      "epoch": 210.42,
+      "grad_norm": 20.15703773498535,
+      "learning_rate": 2.9859859859859863e-06,
+      "loss": 0.2951,
+      "step": 70070
+    },
+    {
+      "epoch": 210.45,
+      "grad_norm": 23.661806106567383,
+      "learning_rate": 2.9849849849849854e-06,
+      "loss": 0.3352,
+      "step": 70080
+    },
+    {
+      "epoch": 210.48,
+      "grad_norm": 14.652510643005371,
+      "learning_rate": 2.983983983983984e-06,
+      "loss": 0.3638,
+      "step": 70090
+    },
+    {
+      "epoch": 210.51,
+      "grad_norm": 14.356799125671387,
+      "learning_rate": 2.982982982982983e-06,
+      "loss": 0.3726,
+      "step": 70100
+    },
+    {
+      "epoch": 210.54,
+      "grad_norm": 9.992854118347168,
+      "learning_rate": 2.9819819819819824e-06,
+      "loss": 0.3373,
+      "step": 70110
+    },
+    {
+      "epoch": 210.57,
+      "grad_norm": 17.905513763427734,
+      "learning_rate": 2.9809809809809814e-06,
+      "loss": 0.3985,
+      "step": 70120
+    },
+    {
+      "epoch": 210.6,
+      "grad_norm": 11.928434371948242,
+      "learning_rate": 2.97997997997998e-06,
+      "loss": 0.3683,
+      "step": 70130
+    },
+    {
+      "epoch": 210.63,
+      "grad_norm": 14.482915878295898,
+      "learning_rate": 2.978978978978979e-06,
+      "loss": 0.3486,
+      "step": 70140
+    },
+    {
+      "epoch": 210.66,
+      "grad_norm": 24.649749755859375,
+      "learning_rate": 2.9779779779779785e-06,
+      "loss": 0.3669,
+      "step": 70150
+    },
+    {
+      "epoch": 210.69,
+      "grad_norm": 29.0954647064209,
+      "learning_rate": 2.9769769769769775e-06,
+      "loss": 0.3762,
+      "step": 70160
+    },
+    {
+      "epoch": 210.72,
+      "grad_norm": 13.338641166687012,
+      "learning_rate": 2.975975975975976e-06,
+      "loss": 0.3237,
+      "step": 70170
+    },
+    {
+      "epoch": 210.75,
+      "grad_norm": 16.26776695251465,
+      "learning_rate": 2.974974974974975e-06,
+      "loss": 0.402,
+      "step": 70180
+    },
+    {
+      "epoch": 210.78,
+      "grad_norm": 12.364664077758789,
+      "learning_rate": 2.973973973973974e-06,
+      "loss": 0.3463,
+      "step": 70190
+    },
+    {
+      "epoch": 210.81,
+      "grad_norm": 22.431854248046875,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.3357,
+      "step": 70200
+    },
+    {
+      "epoch": 210.84,
+      "grad_norm": 13.628296852111816,
+      "learning_rate": 2.971971971971972e-06,
+      "loss": 0.3272,
+      "step": 70210
+    },
+    {
+      "epoch": 210.87,
+      "grad_norm": 15.312373161315918,
+      "learning_rate": 2.970970970970971e-06,
+      "loss": 0.385,
+      "step": 70220
+    },
+    {
+      "epoch": 210.9,
+      "grad_norm": 19.136106491088867,
+      "learning_rate": 2.96996996996997e-06,
+      "loss": 0.3379,
+      "step": 70230
+    },
+    {
+      "epoch": 210.93,
+      "grad_norm": 11.836151123046875,
+      "learning_rate": 2.9689689689689692e-06,
+      "loss": 0.362,
+      "step": 70240
+    },
+    {
+      "epoch": 210.96,
+      "grad_norm": 10.8658447265625,
+      "learning_rate": 2.967967967967968e-06,
+      "loss": 0.3341,
+      "step": 70250
+    },
+    {
+      "epoch": 210.99,
+      "grad_norm": 15.17088794708252,
+      "learning_rate": 2.9669669669669673e-06,
+      "loss": 0.3549,
+      "step": 70260
+    },
+    {
+      "epoch": 211.0,
+      "eval_accuracy": 0.8634,
+      "eval_loss": 0.5437487959861755,
+      "eval_runtime": 12.7269,
+      "eval_samples_per_second": 785.734,
+      "eval_steps_per_second": 3.143,
+      "step": 70263
+    },
+    {
+      "epoch": 211.02,
+      "grad_norm": 17.934616088867188,
+      "learning_rate": 2.9659659659659663e-06,
+      "loss": 0.3092,
+      "step": 70270
+    },
+    {
+      "epoch": 211.05,
+      "grad_norm": 16.59174156188965,
+      "learning_rate": 2.9649649649649653e-06,
+      "loss": 0.2957,
+      "step": 70280
+    },
+    {
+      "epoch": 211.08,
+      "grad_norm": 14.77035903930664,
+      "learning_rate": 2.9639639639639643e-06,
+      "loss": 0.4269,
+      "step": 70290
+    },
+    {
+      "epoch": 211.11,
+      "grad_norm": 14.768013954162598,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.3768,
+      "step": 70300
+    },
+    {
+      "epoch": 211.14,
+      "grad_norm": 21.82026481628418,
+      "learning_rate": 2.9619619619619623e-06,
+      "loss": 0.424,
+      "step": 70310
+    },
+    {
+      "epoch": 211.17,
+      "grad_norm": 19.69100570678711,
+      "learning_rate": 2.9609609609609614e-06,
+      "loss": 0.3217,
+      "step": 70320
+    },
+    {
+      "epoch": 211.2,
+      "grad_norm": 12.162615776062012,
+      "learning_rate": 2.9599599599599604e-06,
+      "loss": 0.3495,
+      "step": 70330
+    },
+    {
+      "epoch": 211.23,
+      "grad_norm": 13.789562225341797,
+      "learning_rate": 2.958958958958959e-06,
+      "loss": 0.3582,
+      "step": 70340
+    },
+    {
+      "epoch": 211.26,
+      "grad_norm": 13.501482963562012,
+      "learning_rate": 2.957957957957958e-06,
+      "loss": 0.3419,
+      "step": 70350
+    },
+    {
+      "epoch": 211.29,
+      "grad_norm": 22.81545639038086,
+      "learning_rate": 2.9569569569569574e-06,
+      "loss": 0.4122,
+      "step": 70360
+    },
+    {
+      "epoch": 211.32,
+      "grad_norm": 14.781512260437012,
+      "learning_rate": 2.9559559559559565e-06,
+      "loss": 0.3127,
+      "step": 70370
+    },
+    {
+      "epoch": 211.35,
+      "grad_norm": 16.116714477539062,
+      "learning_rate": 2.954954954954955e-06,
+      "loss": 0.2986,
+      "step": 70380
+    },
+    {
+      "epoch": 211.38,
+      "grad_norm": 14.07187557220459,
+      "learning_rate": 2.953953953953954e-06,
+      "loss": 0.3683,
+      "step": 70390
+    },
+    {
+      "epoch": 211.41,
+      "grad_norm": 15.567244529724121,
+      "learning_rate": 2.952952952952953e-06,
+      "loss": 0.343,
+      "step": 70400
+    },
+    {
+      "epoch": 211.44,
+      "grad_norm": 15.83078670501709,
+      "learning_rate": 2.9519519519519525e-06,
+      "loss": 0.3717,
+      "step": 70410
+    },
+    {
+      "epoch": 211.47,
+      "grad_norm": 16.801971435546875,
+      "learning_rate": 2.950950950950951e-06,
+      "loss": 0.4336,
+      "step": 70420
+    },
+    {
+      "epoch": 211.5,
+      "grad_norm": 21.0269775390625,
+      "learning_rate": 2.94994994994995e-06,
+      "loss": 0.3142,
+      "step": 70430
+    },
+    {
+      "epoch": 211.53,
+      "grad_norm": 15.853987693786621,
+      "learning_rate": 2.948948948948949e-06,
+      "loss": 0.3575,
+      "step": 70440
+    },
+    {
+      "epoch": 211.56,
+      "grad_norm": 12.963725090026855,
+      "learning_rate": 2.947947947947948e-06,
+      "loss": 0.3531,
+      "step": 70450
+    },
+    {
+      "epoch": 211.59,
+      "grad_norm": 18.111276626586914,
+      "learning_rate": 2.9469469469469468e-06,
+      "loss": 0.3926,
+      "step": 70460
+    },
+    {
+      "epoch": 211.62,
+      "grad_norm": 17.178285598754883,
+      "learning_rate": 2.9459459459459462e-06,
+      "loss": 0.3644,
+      "step": 70470
+    },
+    {
+      "epoch": 211.65,
+      "grad_norm": 14.219483375549316,
+      "learning_rate": 2.9449449449449452e-06,
+      "loss": 0.355,
+      "step": 70480
+    },
+    {
+      "epoch": 211.68,
+      "grad_norm": 12.506941795349121,
+      "learning_rate": 2.9439439439439443e-06,
+      "loss": 0.3305,
+      "step": 70490
+    },
+    {
+      "epoch": 211.71,
+      "grad_norm": 15.919330596923828,
+      "learning_rate": 2.942942942942943e-06,
+      "loss": 0.3271,
+      "step": 70500
+    },
+    {
+      "epoch": 211.74,
+      "grad_norm": 15.00265884399414,
+      "learning_rate": 2.941941941941942e-06,
+      "loss": 0.3289,
+      "step": 70510
+    },
+    {
+      "epoch": 211.77,
+      "grad_norm": 22.43598747253418,
+      "learning_rate": 2.9409409409409413e-06,
+      "loss": 0.3491,
+      "step": 70520
+    },
+    {
+      "epoch": 211.8,
+      "grad_norm": 15.50691032409668,
+      "learning_rate": 2.9399399399399403e-06,
+      "loss": 0.2843,
+      "step": 70530
+    },
+    {
+      "epoch": 211.83,
+      "grad_norm": 15.549955368041992,
+      "learning_rate": 2.9389389389389393e-06,
+      "loss": 0.3622,
+      "step": 70540
+    },
+    {
+      "epoch": 211.86,
+      "grad_norm": 11.739839553833008,
+      "learning_rate": 2.937937937937938e-06,
+      "loss": 0.318,
+      "step": 70550
+    },
+    {
+      "epoch": 211.89,
+      "grad_norm": 13.426733016967773,
+      "learning_rate": 2.936936936936937e-06,
+      "loss": 0.3403,
+      "step": 70560
+    },
+    {
+      "epoch": 211.92,
+      "grad_norm": 11.692191123962402,
+      "learning_rate": 2.9359359359359364e-06,
+      "loss": 0.3824,
+      "step": 70570
+    },
+    {
+      "epoch": 211.95,
+      "grad_norm": 13.062559127807617,
+      "learning_rate": 2.9349349349349354e-06,
+      "loss": 0.3586,
+      "step": 70580
+    },
+    {
+      "epoch": 211.98,
+      "grad_norm": 16.63199234008789,
+      "learning_rate": 2.933933933933934e-06,
+      "loss": 0.3296,
+      "step": 70590
+    },
+    {
+      "epoch": 212.0,
+      "eval_accuracy": 0.8627,
+      "eval_loss": 0.548616886138916,
+      "eval_runtime": 12.6091,
+      "eval_samples_per_second": 793.079,
+      "eval_steps_per_second": 3.172,
+      "step": 70596
+    },
+    {
+      "epoch": 212.01,
+      "grad_norm": 19.16742706298828,
+      "learning_rate": 2.932932932932933e-06,
+      "loss": 0.3831,
+      "step": 70600
+    },
+    {
+      "epoch": 212.04,
+      "grad_norm": 18.141611099243164,
+      "learning_rate": 2.931931931931932e-06,
+      "loss": 0.3623,
+      "step": 70610
+    },
+    {
+      "epoch": 212.07,
+      "grad_norm": 13.375079154968262,
+      "learning_rate": 2.9309309309309315e-06,
+      "loss": 0.3424,
+      "step": 70620
+    },
+    {
+      "epoch": 212.1,
+      "grad_norm": 15.615280151367188,
+      "learning_rate": 2.92992992992993e-06,
+      "loss": 0.3582,
+      "step": 70630
+    },
+    {
+      "epoch": 212.13,
+      "grad_norm": 18.279457092285156,
+      "learning_rate": 2.928928928928929e-06,
+      "loss": 0.3457,
+      "step": 70640
+    },
+    {
+      "epoch": 212.16,
+      "grad_norm": 14.19788646697998,
+      "learning_rate": 2.927927927927928e-06,
+      "loss": 0.3999,
+      "step": 70650
+    },
+    {
+      "epoch": 212.19,
+      "grad_norm": 17.160329818725586,
+      "learning_rate": 2.9269269269269276e-06,
+      "loss": 0.35,
+      "step": 70660
+    },
+    {
+      "epoch": 212.22,
+      "grad_norm": 16.80794906616211,
+      "learning_rate": 2.9259259259259257e-06,
+      "loss": 0.3806,
+      "step": 70670
+    },
+    {
+      "epoch": 212.25,
+      "grad_norm": 14.358774185180664,
+      "learning_rate": 2.924924924924925e-06,
+      "loss": 0.3584,
+      "step": 70680
+    },
+    {
+      "epoch": 212.28,
+      "grad_norm": 15.900832176208496,
+      "learning_rate": 2.923923923923924e-06,
+      "loss": 0.3244,
+      "step": 70690
+    },
+    {
+      "epoch": 212.31,
+      "grad_norm": 11.739288330078125,
+      "learning_rate": 2.9229229229229232e-06,
+      "loss": 0.3769,
+      "step": 70700
+    },
+    {
+      "epoch": 212.34,
+      "grad_norm": 12.818599700927734,
+      "learning_rate": 2.921921921921922e-06,
+      "loss": 0.3405,
+      "step": 70710
+    },
+    {
+      "epoch": 212.37,
+      "grad_norm": 20.68190574645996,
+      "learning_rate": 2.9209209209209213e-06,
+      "loss": 0.3138,
+      "step": 70720
+    },
+    {
+      "epoch": 212.4,
+      "grad_norm": 15.093490600585938,
+      "learning_rate": 2.9199199199199203e-06,
+      "loss": 0.3025,
+      "step": 70730
+    },
+    {
+      "epoch": 212.43,
+      "grad_norm": 20.585172653198242,
+      "learning_rate": 2.9189189189189193e-06,
+      "loss": 0.3515,
+      "step": 70740
+    },
+    {
+      "epoch": 212.46,
+      "grad_norm": 14.319665908813477,
+      "learning_rate": 2.917917917917918e-06,
+      "loss": 0.3655,
+      "step": 70750
+    },
+    {
+      "epoch": 212.49,
+      "grad_norm": 18.768383026123047,
+      "learning_rate": 2.916916916916917e-06,
+      "loss": 0.3772,
+      "step": 70760
+    },
+    {
+      "epoch": 212.52,
+      "grad_norm": 14.011200904846191,
+      "learning_rate": 2.9159159159159163e-06,
+      "loss": 0.35,
+      "step": 70770
+    },
+    {
+      "epoch": 212.55,
+      "grad_norm": 15.779001235961914,
+      "learning_rate": 2.9149149149149154e-06,
+      "loss": 0.3867,
+      "step": 70780
+    },
+    {
+      "epoch": 212.58,
+      "grad_norm": 20.146778106689453,
+      "learning_rate": 2.9139139139139144e-06,
+      "loss": 0.3534,
+      "step": 70790
+    },
+    {
+      "epoch": 212.61,
+      "grad_norm": 16.90038299560547,
+      "learning_rate": 2.912912912912913e-06,
+      "loss": 0.3226,
+      "step": 70800
+    },
+    {
+      "epoch": 212.64,
+      "grad_norm": 32.384342193603516,
+      "learning_rate": 2.911911911911912e-06,
+      "loss": 0.3531,
+      "step": 70810
+    },
+    {
+      "epoch": 212.67,
+      "grad_norm": 10.790935516357422,
+      "learning_rate": 2.9109109109109114e-06,
+      "loss": 0.3363,
+      "step": 70820
+    },
+    {
+      "epoch": 212.7,
+      "grad_norm": 18.093276977539062,
+      "learning_rate": 2.9099099099099105e-06,
+      "loss": 0.3618,
+      "step": 70830
+    },
+    {
+      "epoch": 212.73,
+      "grad_norm": 16.492809295654297,
+      "learning_rate": 2.908908908908909e-06,
+      "loss": 0.3652,
+      "step": 70840
+    },
+    {
+      "epoch": 212.76,
+      "grad_norm": 20.484512329101562,
+      "learning_rate": 2.907907907907908e-06,
+      "loss": 0.405,
+      "step": 70850
+    },
+    {
+      "epoch": 212.79,
+      "grad_norm": 13.032163619995117,
+      "learning_rate": 2.906906906906907e-06,
+      "loss": 0.3354,
+      "step": 70860
+    },
+    {
+      "epoch": 212.82,
+      "grad_norm": 16.742982864379883,
+      "learning_rate": 2.9059059059059065e-06,
+      "loss": 0.3735,
+      "step": 70870
+    },
+    {
+      "epoch": 212.85,
+      "grad_norm": 16.86671257019043,
+      "learning_rate": 2.904904904904905e-06,
+      "loss": 0.3661,
+      "step": 70880
+    },
+    {
+      "epoch": 212.88,
+      "grad_norm": 22.661561965942383,
+      "learning_rate": 2.903903903903904e-06,
+      "loss": 0.3795,
+      "step": 70890
+    },
+    {
+      "epoch": 212.91,
+      "grad_norm": 15.623591423034668,
+      "learning_rate": 2.902902902902903e-06,
+      "loss": 0.3192,
+      "step": 70900
+    },
+    {
+      "epoch": 212.94,
+      "grad_norm": 11.739555358886719,
+      "learning_rate": 2.901901901901902e-06,
+      "loss": 0.3179,
+      "step": 70910
+    },
+    {
+      "epoch": 212.97,
+      "grad_norm": 17.843769073486328,
+      "learning_rate": 2.9009009009009008e-06,
+      "loss": 0.3461,
+      "step": 70920
+    },
+    {
+      "epoch": 213.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5414479374885559,
+      "eval_runtime": 12.9314,
+      "eval_samples_per_second": 773.311,
+      "eval_steps_per_second": 3.093,
+      "step": 70929
+    },
+    {
+      "epoch": 213.0,
+      "grad_norm": 15.815278053283691,
+      "learning_rate": 2.8998998998999002e-06,
+      "loss": 0.2768,
+      "step": 70930
+    },
+    {
+      "epoch": 213.03,
+      "grad_norm": 18.285457611083984,
+      "learning_rate": 2.8988988988988992e-06,
+      "loss": 0.3752,
+      "step": 70940
+    },
+    {
+      "epoch": 213.06,
+      "grad_norm": 20.083330154418945,
+      "learning_rate": 2.8978978978978983e-06,
+      "loss": 0.3914,
+      "step": 70950
+    },
+    {
+      "epoch": 213.09,
+      "grad_norm": 10.061806678771973,
+      "learning_rate": 2.896896896896897e-06,
+      "loss": 0.388,
+      "step": 70960
+    },
+    {
+      "epoch": 213.12,
+      "grad_norm": 17.85187530517578,
+      "learning_rate": 2.895895895895896e-06,
+      "loss": 0.3133,
+      "step": 70970
+    },
+    {
+      "epoch": 213.15,
+      "grad_norm": 15.848522186279297,
+      "learning_rate": 2.8948948948948953e-06,
+      "loss": 0.3217,
+      "step": 70980
+    },
+    {
+      "epoch": 213.18,
+      "grad_norm": 18.622596740722656,
+      "learning_rate": 2.8938938938938943e-06,
+      "loss": 0.343,
+      "step": 70990
+    },
+    {
+      "epoch": 213.21,
+      "grad_norm": 9.936751365661621,
+      "learning_rate": 2.892892892892893e-06,
+      "loss": 0.3122,
+      "step": 71000
+    },
+    {
+      "epoch": 213.24,
+      "grad_norm": 13.649040222167969,
+      "learning_rate": 2.891891891891892e-06,
+      "loss": 0.3097,
+      "step": 71010
+    },
+    {
+      "epoch": 213.27,
+      "grad_norm": 18.253429412841797,
+      "learning_rate": 2.890890890890891e-06,
+      "loss": 0.3606,
+      "step": 71020
+    },
+    {
+      "epoch": 213.3,
+      "grad_norm": 26.373998641967773,
+      "learning_rate": 2.8898898898898904e-06,
+      "loss": 0.3896,
+      "step": 71030
+    },
+    {
+      "epoch": 213.33,
+      "grad_norm": 18.949542999267578,
+      "learning_rate": 2.888888888888889e-06,
+      "loss": 0.3447,
+      "step": 71040
+    },
+    {
+      "epoch": 213.36,
+      "grad_norm": 14.866279602050781,
+      "learning_rate": 2.887887887887888e-06,
+      "loss": 0.3418,
+      "step": 71050
+    },
+    {
+      "epoch": 213.39,
+      "grad_norm": 17.10111427307129,
+      "learning_rate": 2.886886886886887e-06,
+      "loss": 0.3421,
+      "step": 71060
+    },
+    {
+      "epoch": 213.42,
+      "grad_norm": 19.39784812927246,
+      "learning_rate": 2.885885885885886e-06,
+      "loss": 0.3703,
+      "step": 71070
+    },
+    {
+      "epoch": 213.45,
+      "grad_norm": 12.481678009033203,
+      "learning_rate": 2.8848848848848855e-06,
+      "loss": 0.3552,
+      "step": 71080
+    },
+    {
+      "epoch": 213.48,
+      "grad_norm": 21.528884887695312,
+      "learning_rate": 2.883883883883884e-06,
+      "loss": 0.3138,
+      "step": 71090
+    },
+    {
+      "epoch": 213.51,
+      "grad_norm": 14.8321533203125,
+      "learning_rate": 2.882882882882883e-06,
+      "loss": 0.3194,
+      "step": 71100
+    },
+    {
+      "epoch": 213.54,
+      "grad_norm": 23.769821166992188,
+      "learning_rate": 2.881881881881882e-06,
+      "loss": 0.312,
+      "step": 71110
+    },
+    {
+      "epoch": 213.57,
+      "grad_norm": 19.224510192871094,
+      "learning_rate": 2.8808808808808816e-06,
+      "loss": 0.3221,
+      "step": 71120
+    },
+    {
+      "epoch": 213.6,
+      "grad_norm": 21.414955139160156,
+      "learning_rate": 2.8798798798798797e-06,
+      "loss": 0.3841,
+      "step": 71130
+    },
+    {
+      "epoch": 213.63,
+      "grad_norm": 17.994638442993164,
+      "learning_rate": 2.878878878878879e-06,
+      "loss": 0.349,
+      "step": 71140
+    },
+    {
+      "epoch": 213.66,
+      "grad_norm": 10.139829635620117,
+      "learning_rate": 2.877877877877878e-06,
+      "loss": 0.3708,
+      "step": 71150
+    },
+    {
+      "epoch": 213.69,
+      "grad_norm": 23.963756561279297,
+      "learning_rate": 2.8768768768768772e-06,
+      "loss": 0.3476,
+      "step": 71160
+    },
+    {
+      "epoch": 213.72,
+      "grad_norm": 16.0512752532959,
+      "learning_rate": 2.875875875875876e-06,
+      "loss": 0.3313,
+      "step": 71170
+    },
+    {
+      "epoch": 213.75,
+      "grad_norm": 17.633895874023438,
+      "learning_rate": 2.8748748748748753e-06,
+      "loss": 0.3488,
+      "step": 71180
+    },
+    {
+      "epoch": 213.78,
+      "grad_norm": 12.255704879760742,
+      "learning_rate": 2.8738738738738743e-06,
+      "loss": 0.3447,
+      "step": 71190
+    },
+    {
+      "epoch": 213.81,
+      "grad_norm": 11.447075843811035,
+      "learning_rate": 2.8728728728728733e-06,
+      "loss": 0.3144,
+      "step": 71200
+    },
+    {
+      "epoch": 213.84,
+      "grad_norm": 14.498590469360352,
+      "learning_rate": 2.871871871871872e-06,
+      "loss": 0.3781,
+      "step": 71210
+    },
+    {
+      "epoch": 213.87,
+      "grad_norm": 17.893388748168945,
+      "learning_rate": 2.870870870870871e-06,
+      "loss": 0.3686,
+      "step": 71220
+    },
+    {
+      "epoch": 213.9,
+      "grad_norm": 15.591567993164062,
+      "learning_rate": 2.8698698698698703e-06,
+      "loss": 0.3382,
+      "step": 71230
+    },
+    {
+      "epoch": 213.93,
+      "grad_norm": 9.984883308410645,
+      "learning_rate": 2.8688688688688694e-06,
+      "loss": 0.3675,
+      "step": 71240
+    },
+    {
+      "epoch": 213.96,
+      "grad_norm": 25.329174041748047,
+      "learning_rate": 2.867867867867868e-06,
+      "loss": 0.3561,
+      "step": 71250
+    },
+    {
+      "epoch": 213.99,
+      "grad_norm": 10.284937858581543,
+      "learning_rate": 2.866866866866867e-06,
+      "loss": 0.3292,
+      "step": 71260
+    },
+    {
+      "epoch": 214.0,
+      "eval_accuracy": 0.864,
+      "eval_loss": 0.5445184707641602,
+      "eval_runtime": 12.7431,
+      "eval_samples_per_second": 784.739,
+      "eval_steps_per_second": 3.139,
+      "step": 71262
+    },
+    {
+      "epoch": 214.02,
+      "grad_norm": 20.462926864624023,
+      "learning_rate": 2.865865865865866e-06,
+      "loss": 0.3542,
+      "step": 71270
+    },
+    {
+      "epoch": 214.05,
+      "grad_norm": 20.14020347595215,
+      "learning_rate": 2.8648648648648654e-06,
+      "loss": 0.3525,
+      "step": 71280
+    },
+    {
+      "epoch": 214.08,
+      "grad_norm": 16.80267906188965,
+      "learning_rate": 2.863863863863864e-06,
+      "loss": 0.3703,
+      "step": 71290
+    },
+    {
+      "epoch": 214.11,
+      "grad_norm": 14.849209785461426,
+      "learning_rate": 2.862862862862863e-06,
+      "loss": 0.3437,
+      "step": 71300
+    },
+    {
+      "epoch": 214.14,
+      "grad_norm": 14.600312232971191,
+      "learning_rate": 2.861861861861862e-06,
+      "loss": 0.3512,
+      "step": 71310
+    },
+    {
+      "epoch": 214.17,
+      "grad_norm": 26.155092239379883,
+      "learning_rate": 2.860860860860861e-06,
+      "loss": 0.357,
+      "step": 71320
+    },
+    {
+      "epoch": 214.2,
+      "grad_norm": 13.559340476989746,
+      "learning_rate": 2.8598598598598605e-06,
+      "loss": 0.3418,
+      "step": 71330
+    },
+    {
+      "epoch": 214.23,
+      "grad_norm": 22.93210220336914,
+      "learning_rate": 2.858858858858859e-06,
+      "loss": 0.3799,
+      "step": 71340
+    },
+    {
+      "epoch": 214.26,
+      "grad_norm": 23.057954788208008,
+      "learning_rate": 2.857857857857858e-06,
+      "loss": 0.3277,
+      "step": 71350
+    },
+    {
+      "epoch": 214.29,
+      "grad_norm": 13.21805477142334,
+      "learning_rate": 2.856856856856857e-06,
+      "loss": 0.35,
+      "step": 71360
+    },
+    {
+      "epoch": 214.32,
+      "grad_norm": 17.931684494018555,
+      "learning_rate": 2.855855855855856e-06,
+      "loss": 0.3784,
+      "step": 71370
+    },
+    {
+      "epoch": 214.35,
+      "grad_norm": 24.543209075927734,
+      "learning_rate": 2.8548548548548548e-06,
+      "loss": 0.3871,
+      "step": 71380
+    },
+    {
+      "epoch": 214.38,
+      "grad_norm": 13.618919372558594,
+      "learning_rate": 2.853853853853854e-06,
+      "loss": 0.3897,
+      "step": 71390
+    },
+    {
+      "epoch": 214.41,
+      "grad_norm": 16.500347137451172,
+      "learning_rate": 2.8528528528528532e-06,
+      "loss": 0.3233,
+      "step": 71400
+    },
+    {
+      "epoch": 214.44,
+      "grad_norm": 13.615960121154785,
+      "learning_rate": 2.8518518518518522e-06,
+      "loss": 0.3298,
+      "step": 71410
+    },
+    {
+      "epoch": 214.47,
+      "grad_norm": 14.657913208007812,
+      "learning_rate": 2.850850850850851e-06,
+      "loss": 0.3591,
+      "step": 71420
+    },
+    {
+      "epoch": 214.5,
+      "grad_norm": 12.675948143005371,
+      "learning_rate": 2.84984984984985e-06,
+      "loss": 0.3531,
+      "step": 71430
+    },
+    {
+      "epoch": 214.53,
+      "grad_norm": 27.806045532226562,
+      "learning_rate": 2.8488488488488493e-06,
+      "loss": 0.3727,
+      "step": 71440
+    },
+    {
+      "epoch": 214.56,
+      "grad_norm": 20.98536491394043,
+      "learning_rate": 2.8478478478478483e-06,
+      "loss": 0.3847,
+      "step": 71450
+    },
+    {
+      "epoch": 214.59,
+      "grad_norm": 16.01115608215332,
+      "learning_rate": 2.846846846846847e-06,
+      "loss": 0.3896,
+      "step": 71460
+    },
+    {
+      "epoch": 214.62,
+      "grad_norm": 11.670341491699219,
+      "learning_rate": 2.845845845845846e-06,
+      "loss": 0.3713,
+      "step": 71470
+    },
+    {
+      "epoch": 214.65,
+      "grad_norm": 13.51855182647705,
+      "learning_rate": 2.844844844844845e-06,
+      "loss": 0.3737,
+      "step": 71480
+    },
+    {
+      "epoch": 214.68,
+      "grad_norm": 19.568395614624023,
+      "learning_rate": 2.8438438438438444e-06,
+      "loss": 0.3409,
+      "step": 71490
+    },
+    {
+      "epoch": 214.71,
+      "grad_norm": 31.827756881713867,
+      "learning_rate": 2.842842842842843e-06,
+      "loss": 0.2896,
+      "step": 71500
+    },
+    {
+      "epoch": 214.74,
+      "grad_norm": 16.107145309448242,
+      "learning_rate": 2.841841841841842e-06,
+      "loss": 0.3208,
+      "step": 71510
+    },
+    {
+      "epoch": 214.77,
+      "grad_norm": 19.656789779663086,
+      "learning_rate": 2.840840840840841e-06,
+      "loss": 0.3513,
+      "step": 71520
+    },
+    {
+      "epoch": 214.8,
+      "grad_norm": 13.709030151367188,
+      "learning_rate": 2.83983983983984e-06,
+      "loss": 0.3112,
+      "step": 71530
+    },
+    {
+      "epoch": 214.83,
+      "grad_norm": 21.88645362854004,
+      "learning_rate": 2.8388388388388386e-06,
+      "loss": 0.3697,
+      "step": 71540
+    },
+    {
+      "epoch": 214.86,
+      "grad_norm": 8.816058158874512,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.3092,
+      "step": 71550
+    },
+    {
+      "epoch": 214.89,
+      "grad_norm": 18.335590362548828,
+      "learning_rate": 2.836836836836837e-06,
+      "loss": 0.3109,
+      "step": 71560
+    },
+    {
+      "epoch": 214.92,
+      "grad_norm": 19.70177459716797,
+      "learning_rate": 2.835835835835836e-06,
+      "loss": 0.3615,
+      "step": 71570
+    },
+    {
+      "epoch": 214.95,
+      "grad_norm": 15.163211822509766,
+      "learning_rate": 2.8348348348348347e-06,
+      "loss": 0.3487,
+      "step": 71580
+    },
+    {
+      "epoch": 214.98,
+      "grad_norm": 14.805184364318848,
+      "learning_rate": 2.8338338338338337e-06,
+      "loss": 0.3622,
+      "step": 71590
+    },
+    {
+      "epoch": 215.0,
+      "eval_accuracy": 0.8626,
+      "eval_loss": 0.5437569618225098,
+      "eval_runtime": 12.6888,
+      "eval_samples_per_second": 788.094,
+      "eval_steps_per_second": 3.152,
+      "step": 71595
+    },
+    {
+      "epoch": 215.02,
+      "grad_norm": 12.247604370117188,
+      "learning_rate": 2.832832832832833e-06,
+      "loss": 0.2893,
+      "step": 71600
+    },
+    {
+      "epoch": 215.05,
+      "grad_norm": 16.630338668823242,
+      "learning_rate": 2.831831831831832e-06,
+      "loss": 0.3693,
+      "step": 71610
+    },
+    {
+      "epoch": 215.08,
+      "grad_norm": 15.363871574401855,
+      "learning_rate": 2.830830830830831e-06,
+      "loss": 0.3134,
+      "step": 71620
+    },
+    {
+      "epoch": 215.11,
+      "grad_norm": 22.891233444213867,
+      "learning_rate": 2.82982982982983e-06,
+      "loss": 0.3665,
+      "step": 71630
+    },
+    {
+      "epoch": 215.14,
+      "grad_norm": 16.095914840698242,
+      "learning_rate": 2.828828828828829e-06,
+      "loss": 0.3664,
+      "step": 71640
+    },
+    {
+      "epoch": 215.17,
+      "grad_norm": 12.877725601196289,
+      "learning_rate": 2.8278278278278283e-06,
+      "loss": 0.3622,
+      "step": 71650
+    },
+    {
+      "epoch": 215.2,
+      "grad_norm": 12.117210388183594,
+      "learning_rate": 2.8268268268268273e-06,
+      "loss": 0.3454,
+      "step": 71660
+    },
+    {
+      "epoch": 215.23,
+      "grad_norm": 13.582042694091797,
+      "learning_rate": 2.825825825825826e-06,
+      "loss": 0.3816,
+      "step": 71670
+    },
+    {
+      "epoch": 215.26,
+      "grad_norm": 13.540081977844238,
+      "learning_rate": 2.824824824824825e-06,
+      "loss": 0.321,
+      "step": 71680
+    },
+    {
+      "epoch": 215.29,
+      "grad_norm": 15.139362335205078,
+      "learning_rate": 2.8238238238238243e-06,
+      "loss": 0.3202,
+      "step": 71690
+    },
+    {
+      "epoch": 215.32,
+      "grad_norm": 17.130664825439453,
+      "learning_rate": 2.8228228228228234e-06,
+      "loss": 0.3843,
+      "step": 71700
+    },
+    {
+      "epoch": 215.35,
+      "grad_norm": 17.52095603942871,
+      "learning_rate": 2.821821821821822e-06,
+      "loss": 0.3649,
+      "step": 71710
+    },
+    {
+      "epoch": 215.38,
+      "grad_norm": 17.579784393310547,
+      "learning_rate": 2.820820820820821e-06,
+      "loss": 0.4044,
+      "step": 71720
+    },
+    {
+      "epoch": 215.41,
+      "grad_norm": 17.84904670715332,
+      "learning_rate": 2.81981981981982e-06,
+      "loss": 0.3132,
+      "step": 71730
+    },
+    {
+      "epoch": 215.44,
+      "grad_norm": 15.848404884338379,
+      "learning_rate": 2.8188188188188194e-06,
+      "loss": 0.3469,
+      "step": 71740
+    },
+    {
+      "epoch": 215.47,
+      "grad_norm": 14.817167282104492,
+      "learning_rate": 2.817817817817818e-06,
+      "loss": 0.3274,
+      "step": 71750
+    },
+    {
+      "epoch": 215.5,
+      "grad_norm": 15.88916015625,
+      "learning_rate": 2.816816816816817e-06,
+      "loss": 0.4052,
+      "step": 71760
+    },
+    {
+      "epoch": 215.53,
+      "grad_norm": 13.936515808105469,
+      "learning_rate": 2.815815815815816e-06,
+      "loss": 0.3755,
+      "step": 71770
+    },
+    {
+      "epoch": 215.56,
+      "grad_norm": 15.714937210083008,
+      "learning_rate": 2.814814814814815e-06,
+      "loss": 0.3245,
+      "step": 71780
+    },
+    {
+      "epoch": 215.59,
+      "grad_norm": 20.432342529296875,
+      "learning_rate": 2.8138138138138137e-06,
+      "loss": 0.3712,
+      "step": 71790
+    },
+    {
+      "epoch": 215.62,
+      "grad_norm": 10.723136901855469,
+      "learning_rate": 2.812812812812813e-06,
+      "loss": 0.3476,
+      "step": 71800
+    },
+    {
+      "epoch": 215.65,
+      "grad_norm": 17.66670036315918,
+      "learning_rate": 2.811811811811812e-06,
+      "loss": 0.3201,
+      "step": 71810
+    },
+    {
+      "epoch": 215.68,
+      "grad_norm": 18.749902725219727,
+      "learning_rate": 2.810810810810811e-06,
+      "loss": 0.3556,
+      "step": 71820
+    },
+    {
+      "epoch": 215.71,
+      "grad_norm": 16.695878982543945,
+      "learning_rate": 2.8098098098098097e-06,
+      "loss": 0.351,
+      "step": 71830
+    },
+    {
+      "epoch": 215.74,
+      "grad_norm": 17.58240509033203,
+      "learning_rate": 2.8088088088088088e-06,
+      "loss": 0.335,
+      "step": 71840
+    },
+    {
+      "epoch": 215.77,
+      "grad_norm": 12.699224472045898,
+      "learning_rate": 2.807807807807808e-06,
+      "loss": 0.2976,
+      "step": 71850
+    },
+    {
+      "epoch": 215.8,
+      "grad_norm": 14.48170280456543,
+      "learning_rate": 2.8068068068068072e-06,
+      "loss": 0.3236,
+      "step": 71860
+    },
+    {
+      "epoch": 215.83,
+      "grad_norm": 19.756351470947266,
+      "learning_rate": 2.8058058058058062e-06,
+      "loss": 0.4247,
+      "step": 71870
+    },
+    {
+      "epoch": 215.86,
+      "grad_norm": 15.066643714904785,
+      "learning_rate": 2.804804804804805e-06,
+      "loss": 0.3018,
+      "step": 71880
+    },
+    {
+      "epoch": 215.89,
+      "grad_norm": 13.934998512268066,
+      "learning_rate": 2.803803803803804e-06,
+      "loss": 0.353,
+      "step": 71890
+    },
+    {
+      "epoch": 215.92,
+      "grad_norm": 15.664441108703613,
+      "learning_rate": 2.8028028028028033e-06,
+      "loss": 0.364,
+      "step": 71900
+    },
+    {
+      "epoch": 215.95,
+      "grad_norm": 16.49651336669922,
+      "learning_rate": 2.8018018018018023e-06,
+      "loss": 0.3617,
+      "step": 71910
+    },
+    {
+      "epoch": 215.98,
+      "grad_norm": 15.116989135742188,
+      "learning_rate": 2.800800800800801e-06,
+      "loss": 0.3724,
+      "step": 71920
+    },
+    {
+      "epoch": 216.0,
+      "eval_accuracy": 0.8665,
+      "eval_loss": 0.5359203219413757,
+      "eval_runtime": 12.9755,
+      "eval_samples_per_second": 770.681,
+      "eval_steps_per_second": 3.083,
+      "step": 71928
+    },
+    {
+      "epoch": 216.01,
+      "grad_norm": 15.53590202331543,
+      "learning_rate": 2.7997997997998e-06,
+      "loss": 0.3187,
+      "step": 71930
+    },
+    {
+      "epoch": 216.04,
+      "grad_norm": 13.980267524719238,
+      "learning_rate": 2.798798798798799e-06,
+      "loss": 0.339,
+      "step": 71940
+    },
+    {
+      "epoch": 216.07,
+      "grad_norm": 12.184996604919434,
+      "learning_rate": 2.7977977977977984e-06,
+      "loss": 0.3321,
+      "step": 71950
+    },
+    {
+      "epoch": 216.1,
+      "grad_norm": 18.923675537109375,
+      "learning_rate": 2.796796796796797e-06,
+      "loss": 0.3251,
+      "step": 71960
+    },
+    {
+      "epoch": 216.13,
+      "grad_norm": 17.231374740600586,
+      "learning_rate": 2.795795795795796e-06,
+      "loss": 0.3732,
+      "step": 71970
+    },
+    {
+      "epoch": 216.16,
+      "grad_norm": 19.041086196899414,
+      "learning_rate": 2.794794794794795e-06,
+      "loss": 0.3657,
+      "step": 71980
+    },
+    {
+      "epoch": 216.19,
+      "grad_norm": 18.589092254638672,
+      "learning_rate": 2.793793793793794e-06,
+      "loss": 0.3617,
+      "step": 71990
+    },
+    {
+      "epoch": 216.22,
+      "grad_norm": 15.953536033630371,
+      "learning_rate": 2.7927927927927926e-06,
+      "loss": 0.3261,
+      "step": 72000
+    },
+    {
+      "epoch": 216.25,
+      "grad_norm": 11.201769828796387,
+      "learning_rate": 2.791791791791792e-06,
+      "loss": 0.336,
+      "step": 72010
+    },
+    {
+      "epoch": 216.28,
+      "grad_norm": 21.10941505432129,
+      "learning_rate": 2.790790790790791e-06,
+      "loss": 0.3414,
+      "step": 72020
+    },
+    {
+      "epoch": 216.31,
+      "grad_norm": 12.811768531799316,
+      "learning_rate": 2.78978978978979e-06,
+      "loss": 0.346,
+      "step": 72030
+    },
+    {
+      "epoch": 216.34,
+      "grad_norm": 21.267967224121094,
+      "learning_rate": 2.7887887887887887e-06,
+      "loss": 0.2901,
+      "step": 72040
+    },
+    {
+      "epoch": 216.37,
+      "grad_norm": 15.422242164611816,
+      "learning_rate": 2.7877877877877877e-06,
+      "loss": 0.3746,
+      "step": 72050
+    },
+    {
+      "epoch": 216.4,
+      "grad_norm": 10.1339750289917,
+      "learning_rate": 2.786786786786787e-06,
+      "loss": 0.3689,
+      "step": 72060
+    },
+    {
+      "epoch": 216.43,
+      "grad_norm": 23.65520477294922,
+      "learning_rate": 2.785785785785786e-06,
+      "loss": 0.3694,
+      "step": 72070
+    },
+    {
+      "epoch": 216.46,
+      "grad_norm": 20.862911224365234,
+      "learning_rate": 2.7847847847847848e-06,
+      "loss": 0.3554,
+      "step": 72080
+    },
+    {
+      "epoch": 216.49,
+      "grad_norm": 18.490406036376953,
+      "learning_rate": 2.783783783783784e-06,
+      "loss": 0.3142,
+      "step": 72090
+    },
+    {
+      "epoch": 216.52,
+      "grad_norm": 22.16808319091797,
+      "learning_rate": 2.782782782782783e-06,
+      "loss": 0.342,
+      "step": 72100
+    },
+    {
+      "epoch": 216.55,
+      "grad_norm": 14.817203521728516,
+      "learning_rate": 2.7817817817817823e-06,
+      "loss": 0.3463,
+      "step": 72110
+    },
+    {
+      "epoch": 216.58,
+      "grad_norm": 17.932903289794922,
+      "learning_rate": 2.780780780780781e-06,
+      "loss": 0.3506,
+      "step": 72120
+    },
+    {
+      "epoch": 216.61,
+      "grad_norm": 18.520526885986328,
+      "learning_rate": 2.77977977977978e-06,
+      "loss": 0.3417,
+      "step": 72130
+    },
+    {
+      "epoch": 216.64,
+      "grad_norm": 21.82206153869629,
+      "learning_rate": 2.778778778778779e-06,
+      "loss": 0.3358,
+      "step": 72140
+    },
+    {
+      "epoch": 216.67,
+      "grad_norm": 18.11307716369629,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.3721,
+      "step": 72150
+    },
+    {
+      "epoch": 216.7,
+      "grad_norm": 12.036067008972168,
+      "learning_rate": 2.7767767767767774e-06,
+      "loss": 0.3315,
+      "step": 72160
+    },
+    {
+      "epoch": 216.73,
+      "grad_norm": 17.150074005126953,
+      "learning_rate": 2.775775775775776e-06,
+      "loss": 0.3279,
+      "step": 72170
+    },
+    {
+      "epoch": 216.76,
+      "grad_norm": 13.77971363067627,
+      "learning_rate": 2.774774774774775e-06,
+      "loss": 0.3187,
+      "step": 72180
+    },
+    {
+      "epoch": 216.79,
+      "grad_norm": 15.712998390197754,
+      "learning_rate": 2.773773773773774e-06,
+      "loss": 0.3354,
+      "step": 72190
+    },
+    {
+      "epoch": 216.82,
+      "grad_norm": 20.30086326599121,
+      "learning_rate": 2.7727727727727734e-06,
+      "loss": 0.3851,
+      "step": 72200
+    },
+    {
+      "epoch": 216.85,
+      "grad_norm": 25.87421417236328,
+      "learning_rate": 2.771771771771772e-06,
+      "loss": 0.3657,
+      "step": 72210
+    },
+    {
+      "epoch": 216.88,
+      "grad_norm": 15.357707023620605,
+      "learning_rate": 2.770770770770771e-06,
+      "loss": 0.3615,
+      "step": 72220
+    },
+    {
+      "epoch": 216.91,
+      "grad_norm": 12.295831680297852,
+      "learning_rate": 2.76976976976977e-06,
+      "loss": 0.3466,
+      "step": 72230
+    },
+    {
+      "epoch": 216.94,
+      "grad_norm": 13.451245307922363,
+      "learning_rate": 2.768768768768769e-06,
+      "loss": 0.3672,
+      "step": 72240
+    },
+    {
+      "epoch": 216.97,
+      "grad_norm": 9.265341758728027,
+      "learning_rate": 2.7677677677677677e-06,
+      "loss": 0.3204,
+      "step": 72250
+    },
+    {
+      "epoch": 217.0,
+      "grad_norm": 16.115886688232422,
+      "learning_rate": 2.766766766766767e-06,
+      "loss": 0.3352,
+      "step": 72260
+    },
+    {
+      "epoch": 217.0,
+      "eval_accuracy": 0.8658,
+      "eval_loss": 0.5410099029541016,
+      "eval_runtime": 12.6545,
+      "eval_samples_per_second": 790.235,
+      "eval_steps_per_second": 3.161,
+      "step": 72261
+    },
+    {
+      "epoch": 217.03,
+      "grad_norm": 15.168283462524414,
+      "learning_rate": 2.765765765765766e-06,
+      "loss": 0.3233,
+      "step": 72270
+    },
+    {
+      "epoch": 217.06,
+      "grad_norm": 20.84393882751465,
+      "learning_rate": 2.764764764764765e-06,
+      "loss": 0.3251,
+      "step": 72280
+    },
+    {
+      "epoch": 217.09,
+      "grad_norm": 14.30004596710205,
+      "learning_rate": 2.7637637637637637e-06,
+      "loss": 0.319,
+      "step": 72290
+    },
+    {
+      "epoch": 217.12,
+      "grad_norm": 27.271883010864258,
+      "learning_rate": 2.7627627627627628e-06,
+      "loss": 0.3411,
+      "step": 72300
+    },
+    {
+      "epoch": 217.15,
+      "grad_norm": 13.627557754516602,
+      "learning_rate": 2.761761761761762e-06,
+      "loss": 0.3557,
+      "step": 72310
+    },
+    {
+      "epoch": 217.18,
+      "grad_norm": 16.206817626953125,
+      "learning_rate": 2.7607607607607612e-06,
+      "loss": 0.3595,
+      "step": 72320
+    },
+    {
+      "epoch": 217.21,
+      "grad_norm": 14.963168144226074,
+      "learning_rate": 2.75975975975976e-06,
+      "loss": 0.3547,
+      "step": 72330
+    },
+    {
+      "epoch": 217.24,
+      "grad_norm": 15.973031997680664,
+      "learning_rate": 2.758758758758759e-06,
+      "loss": 0.3536,
+      "step": 72340
+    },
+    {
+      "epoch": 217.27,
+      "grad_norm": 15.79863166809082,
+      "learning_rate": 2.757757757757758e-06,
+      "loss": 0.3612,
+      "step": 72350
+    },
+    {
+      "epoch": 217.3,
+      "grad_norm": 15.301898956298828,
+      "learning_rate": 2.7567567567567573e-06,
+      "loss": 0.3919,
+      "step": 72360
+    },
+    {
+      "epoch": 217.33,
+      "grad_norm": 18.607952117919922,
+      "learning_rate": 2.755755755755756e-06,
+      "loss": 0.337,
+      "step": 72370
+    },
+    {
+      "epoch": 217.36,
+      "grad_norm": 18.7547607421875,
+      "learning_rate": 2.754754754754755e-06,
+      "loss": 0.3206,
+      "step": 72380
+    },
+    {
+      "epoch": 217.39,
+      "grad_norm": 21.169921875,
+      "learning_rate": 2.753753753753754e-06,
+      "loss": 0.3448,
+      "step": 72390
+    },
+    {
+      "epoch": 217.42,
+      "grad_norm": 9.937636375427246,
+      "learning_rate": 2.752752752752753e-06,
+      "loss": 0.2925,
+      "step": 72400
+    },
+    {
+      "epoch": 217.45,
+      "grad_norm": 16.16466522216797,
+      "learning_rate": 2.7517517517517524e-06,
+      "loss": 0.3696,
+      "step": 72410
+    },
+    {
+      "epoch": 217.48,
+      "grad_norm": 15.18020248413086,
+      "learning_rate": 2.750750750750751e-06,
+      "loss": 0.3594,
+      "step": 72420
+    },
+    {
+      "epoch": 217.51,
+      "grad_norm": 20.718517303466797,
+      "learning_rate": 2.74974974974975e-06,
+      "loss": 0.3423,
+      "step": 72430
+    },
+    {
+      "epoch": 217.54,
+      "grad_norm": 12.042766571044922,
+      "learning_rate": 2.748748748748749e-06,
+      "loss": 0.3246,
+      "step": 72440
+    },
+    {
+      "epoch": 217.57,
+      "grad_norm": 20.97349739074707,
+      "learning_rate": 2.747747747747748e-06,
+      "loss": 0.3283,
+      "step": 72450
+    },
+    {
+      "epoch": 217.6,
+      "grad_norm": 13.060112953186035,
+      "learning_rate": 2.7467467467467466e-06,
+      "loss": 0.3487,
+      "step": 72460
+    },
+    {
+      "epoch": 217.63,
+      "grad_norm": 19.48906135559082,
+      "learning_rate": 2.745745745745746e-06,
+      "loss": 0.3118,
+      "step": 72470
+    },
+    {
+      "epoch": 217.66,
+      "grad_norm": 13.089376449584961,
+      "learning_rate": 2.744744744744745e-06,
+      "loss": 0.3096,
+      "step": 72480
+    },
+    {
+      "epoch": 217.69,
+      "grad_norm": 17.093765258789062,
+      "learning_rate": 2.743743743743744e-06,
+      "loss": 0.355,
+      "step": 72490
+    },
+    {
+      "epoch": 217.72,
+      "grad_norm": 24.91587257385254,
+      "learning_rate": 2.7427427427427427e-06,
+      "loss": 0.36,
+      "step": 72500
+    },
+    {
+      "epoch": 217.75,
+      "grad_norm": 9.432390213012695,
+      "learning_rate": 2.7417417417417417e-06,
+      "loss": 0.3515,
+      "step": 72510
+    },
+    {
+      "epoch": 217.78,
+      "grad_norm": 14.507742881774902,
+      "learning_rate": 2.740740740740741e-06,
+      "loss": 0.3911,
+      "step": 72520
+    },
+    {
+      "epoch": 217.81,
+      "grad_norm": 15.03933334350586,
+      "learning_rate": 2.73973973973974e-06,
+      "loss": 0.3513,
+      "step": 72530
+    },
+    {
+      "epoch": 217.84,
+      "grad_norm": 14.733933448791504,
+      "learning_rate": 2.7387387387387388e-06,
+      "loss": 0.3412,
+      "step": 72540
+    },
+    {
+      "epoch": 217.87,
+      "grad_norm": 21.036224365234375,
+      "learning_rate": 2.737737737737738e-06,
+      "loss": 0.3362,
+      "step": 72550
+    },
+    {
+      "epoch": 217.9,
+      "grad_norm": 19.787033081054688,
+      "learning_rate": 2.736736736736737e-06,
+      "loss": 0.3287,
+      "step": 72560
+    },
+    {
+      "epoch": 217.93,
+      "grad_norm": 14.58094310760498,
+      "learning_rate": 2.7357357357357363e-06,
+      "loss": 0.3498,
+      "step": 72570
+    },
+    {
+      "epoch": 217.96,
+      "grad_norm": 28.4337100982666,
+      "learning_rate": 2.734734734734735e-06,
+      "loss": 0.3444,
+      "step": 72580
+    },
+    {
+      "epoch": 217.99,
+      "grad_norm": 12.295095443725586,
+      "learning_rate": 2.733733733733734e-06,
+      "loss": 0.3484,
+      "step": 72590
+    },
+    {
+      "epoch": 218.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.54071444272995,
+      "eval_runtime": 12.8042,
+      "eval_samples_per_second": 780.991,
+      "eval_steps_per_second": 3.124,
+      "step": 72594
+    },
+    {
+      "epoch": 218.02,
+      "grad_norm": 8.817733764648438,
+      "learning_rate": 2.732732732732733e-06,
+      "loss": 0.4334,
+      "step": 72600
+    },
+    {
+      "epoch": 218.05,
+      "grad_norm": 16.361373901367188,
+      "learning_rate": 2.731731731731732e-06,
+      "loss": 0.3106,
+      "step": 72610
+    },
+    {
+      "epoch": 218.08,
+      "grad_norm": 16.420452117919922,
+      "learning_rate": 2.7307307307307305e-06,
+      "loss": 0.3396,
+      "step": 72620
+    },
+    {
+      "epoch": 218.11,
+      "grad_norm": 17.834789276123047,
+      "learning_rate": 2.72972972972973e-06,
+      "loss": 0.3533,
+      "step": 72630
+    },
+    {
+      "epoch": 218.14,
+      "grad_norm": 16.025741577148438,
+      "learning_rate": 2.728728728728729e-06,
+      "loss": 0.3304,
+      "step": 72640
+    },
+    {
+      "epoch": 218.17,
+      "grad_norm": 19.46750831604004,
+      "learning_rate": 2.727727727727728e-06,
+      "loss": 0.38,
+      "step": 72650
+    },
+    {
+      "epoch": 218.2,
+      "grad_norm": 19.556421279907227,
+      "learning_rate": 2.7267267267267274e-06,
+      "loss": 0.3597,
+      "step": 72660
+    },
+    {
+      "epoch": 218.23,
+      "grad_norm": 18.525861740112305,
+      "learning_rate": 2.7257257257257256e-06,
+      "loss": 0.357,
+      "step": 72670
+    },
+    {
+      "epoch": 218.26,
+      "grad_norm": 18.434587478637695,
+      "learning_rate": 2.724724724724725e-06,
+      "loss": 0.3083,
+      "step": 72680
+    },
+    {
+      "epoch": 218.29,
+      "grad_norm": 16.818334579467773,
+      "learning_rate": 2.723723723723724e-06,
+      "loss": 0.3286,
+      "step": 72690
+    },
+    {
+      "epoch": 218.32,
+      "grad_norm": 26.2186222076416,
+      "learning_rate": 2.722722722722723e-06,
+      "loss": 0.3184,
+      "step": 72700
+    },
+    {
+      "epoch": 218.35,
+      "grad_norm": 14.60560417175293,
+      "learning_rate": 2.7217217217217217e-06,
+      "loss": 0.3429,
+      "step": 72710
+    },
+    {
+      "epoch": 218.38,
+      "grad_norm": 17.65956687927246,
+      "learning_rate": 2.720720720720721e-06,
+      "loss": 0.3557,
+      "step": 72720
+    },
+    {
+      "epoch": 218.41,
+      "grad_norm": 21.668153762817383,
+      "learning_rate": 2.71971971971972e-06,
+      "loss": 0.3525,
+      "step": 72730
+    },
+    {
+      "epoch": 218.44,
+      "grad_norm": 24.232179641723633,
+      "learning_rate": 2.718718718718719e-06,
+      "loss": 0.3602,
+      "step": 72740
+    },
+    {
+      "epoch": 218.47,
+      "grad_norm": 12.232305526733398,
+      "learning_rate": 2.7177177177177177e-06,
+      "loss": 0.3528,
+      "step": 72750
+    },
+    {
+      "epoch": 218.5,
+      "grad_norm": 18.636083602905273,
+      "learning_rate": 2.7167167167167168e-06,
+      "loss": 0.3393,
+      "step": 72760
+    },
+    {
+      "epoch": 218.53,
+      "grad_norm": 17.38655662536621,
+      "learning_rate": 2.715715715715716e-06,
+      "loss": 0.3396,
+      "step": 72770
+    },
+    {
+      "epoch": 218.56,
+      "grad_norm": 15.567644119262695,
+      "learning_rate": 2.7147147147147152e-06,
+      "loss": 0.3312,
+      "step": 72780
+    },
+    {
+      "epoch": 218.59,
+      "grad_norm": 29.496212005615234,
+      "learning_rate": 2.713713713713714e-06,
+      "loss": 0.3613,
+      "step": 72790
+    },
+    {
+      "epoch": 218.62,
+      "grad_norm": 18.225418090820312,
+      "learning_rate": 2.712712712712713e-06,
+      "loss": 0.3551,
+      "step": 72800
+    },
+    {
+      "epoch": 218.65,
+      "grad_norm": 14.601926803588867,
+      "learning_rate": 2.711711711711712e-06,
+      "loss": 0.3559,
+      "step": 72810
+    },
+    {
+      "epoch": 218.68,
+      "grad_norm": 12.434026718139648,
+      "learning_rate": 2.7107107107107113e-06,
+      "loss": 0.3384,
+      "step": 72820
+    },
+    {
+      "epoch": 218.71,
+      "grad_norm": 16.444904327392578,
+      "learning_rate": 2.70970970970971e-06,
+      "loss": 0.3617,
+      "step": 72830
+    },
+    {
+      "epoch": 218.74,
+      "grad_norm": 12.206204414367676,
+      "learning_rate": 2.708708708708709e-06,
+      "loss": 0.3289,
+      "step": 72840
+    },
+    {
+      "epoch": 218.77,
+      "grad_norm": 20.470401763916016,
+      "learning_rate": 2.707707707707708e-06,
+      "loss": 0.3059,
+      "step": 72850
+    },
+    {
+      "epoch": 218.8,
+      "grad_norm": 22.455564498901367,
+      "learning_rate": 2.706706706706707e-06,
+      "loss": 0.3397,
+      "step": 72860
+    },
+    {
+      "epoch": 218.83,
+      "grad_norm": 17.63998031616211,
+      "learning_rate": 2.7057057057057055e-06,
+      "loss": 0.3462,
+      "step": 72870
+    },
+    {
+      "epoch": 218.86,
+      "grad_norm": 19.806703567504883,
+      "learning_rate": 2.704704704704705e-06,
+      "loss": 0.3902,
+      "step": 72880
+    },
+    {
+      "epoch": 218.89,
+      "grad_norm": 14.890491485595703,
+      "learning_rate": 2.703703703703704e-06,
+      "loss": 0.3614,
+      "step": 72890
+    },
+    {
+      "epoch": 218.92,
+      "grad_norm": 18.68931007385254,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.3779,
+      "step": 72900
+    },
+    {
+      "epoch": 218.95,
+      "grad_norm": 27.1975154876709,
+      "learning_rate": 2.7017017017017016e-06,
+      "loss": 0.3676,
+      "step": 72910
+    },
+    {
+      "epoch": 218.98,
+      "grad_norm": 15.131671905517578,
+      "learning_rate": 2.7007007007007006e-06,
+      "loss": 0.3109,
+      "step": 72920
+    },
+    {
+      "epoch": 219.0,
+      "eval_accuracy": 0.8653,
+      "eval_loss": 0.5403647422790527,
+      "eval_runtime": 13.0712,
+      "eval_samples_per_second": 765.039,
+      "eval_steps_per_second": 3.06,
+      "step": 72927
+    },
+    {
+      "epoch": 219.01,
+      "grad_norm": 12.972577095031738,
+      "learning_rate": 2.6996996996997e-06,
+      "loss": 0.3312,
+      "step": 72930
+    },
+    {
+      "epoch": 219.04,
+      "grad_norm": 15.431875228881836,
+      "learning_rate": 2.698698698698699e-06,
+      "loss": 0.323,
+      "step": 72940
+    },
+    {
+      "epoch": 219.07,
+      "grad_norm": 20.598209381103516,
+      "learning_rate": 2.697697697697698e-06,
+      "loss": 0.3541,
+      "step": 72950
+    },
+    {
+      "epoch": 219.1,
+      "grad_norm": 18.611391067504883,
+      "learning_rate": 2.6966966966966967e-06,
+      "loss": 0.3385,
+      "step": 72960
+    },
+    {
+      "epoch": 219.13,
+      "grad_norm": 20.01825523376465,
+      "learning_rate": 2.6956956956956957e-06,
+      "loss": 0.3192,
+      "step": 72970
+    },
+    {
+      "epoch": 219.16,
+      "grad_norm": 18.88039207458496,
+      "learning_rate": 2.694694694694695e-06,
+      "loss": 0.3446,
+      "step": 72980
+    },
+    {
+      "epoch": 219.19,
+      "grad_norm": 13.4527006149292,
+      "learning_rate": 2.693693693693694e-06,
+      "loss": 0.3256,
+      "step": 72990
+    },
+    {
+      "epoch": 219.22,
+      "grad_norm": 14.116559028625488,
+      "learning_rate": 2.6926926926926928e-06,
+      "loss": 0.353,
+      "step": 73000
+    },
+    {
+      "epoch": 219.25,
+      "grad_norm": 10.352441787719727,
+      "learning_rate": 2.691691691691692e-06,
+      "loss": 0.3378,
+      "step": 73010
+    },
+    {
+      "epoch": 219.28,
+      "grad_norm": 18.187171936035156,
+      "learning_rate": 2.690690690690691e-06,
+      "loss": 0.4012,
+      "step": 73020
+    },
+    {
+      "epoch": 219.31,
+      "grad_norm": 13.539438247680664,
+      "learning_rate": 2.6896896896896903e-06,
+      "loss": 0.3255,
+      "step": 73030
+    },
+    {
+      "epoch": 219.34,
+      "grad_norm": 14.488775253295898,
+      "learning_rate": 2.688688688688689e-06,
+      "loss": 0.3194,
+      "step": 73040
+    },
+    {
+      "epoch": 219.37,
+      "grad_norm": 25.33340072631836,
+      "learning_rate": 2.687687687687688e-06,
+      "loss": 0.3364,
+      "step": 73050
+    },
+    {
+      "epoch": 219.4,
+      "grad_norm": 15.877584457397461,
+      "learning_rate": 2.686686686686687e-06,
+      "loss": 0.3299,
+      "step": 73060
+    },
+    {
+      "epoch": 219.43,
+      "grad_norm": 16.926477432250977,
+      "learning_rate": 2.685685685685686e-06,
+      "loss": 0.3277,
+      "step": 73070
+    },
+    {
+      "epoch": 219.46,
+      "grad_norm": 16.368385314941406,
+      "learning_rate": 2.6846846846846845e-06,
+      "loss": 0.3303,
+      "step": 73080
+    },
+    {
+      "epoch": 219.49,
+      "grad_norm": 23.28862762451172,
+      "learning_rate": 2.683683683683684e-06,
+      "loss": 0.3294,
+      "step": 73090
+    },
+    {
+      "epoch": 219.52,
+      "grad_norm": 12.245881080627441,
+      "learning_rate": 2.682682682682683e-06,
+      "loss": 0.3463,
+      "step": 73100
+    },
+    {
+      "epoch": 219.55,
+      "grad_norm": 19.784948348999023,
+      "learning_rate": 2.681681681681682e-06,
+      "loss": 0.3587,
+      "step": 73110
+    },
+    {
+      "epoch": 219.58,
+      "grad_norm": 19.28864288330078,
+      "learning_rate": 2.6806806806806806e-06,
+      "loss": 0.3715,
+      "step": 73120
+    },
+    {
+      "epoch": 219.61,
+      "grad_norm": 19.467126846313477,
+      "learning_rate": 2.6796796796796796e-06,
+      "loss": 0.3605,
+      "step": 73130
+    },
+    {
+      "epoch": 219.64,
+      "grad_norm": 19.61667251586914,
+      "learning_rate": 2.678678678678679e-06,
+      "loss": 0.3324,
+      "step": 73140
+    },
+    {
+      "epoch": 219.67,
+      "grad_norm": 23.469646453857422,
+      "learning_rate": 2.677677677677678e-06,
+      "loss": 0.3486,
+      "step": 73150
+    },
+    {
+      "epoch": 219.7,
+      "grad_norm": 15.14398193359375,
+      "learning_rate": 2.6766766766766766e-06,
+      "loss": 0.3427,
+      "step": 73160
+    },
+    {
+      "epoch": 219.73,
+      "grad_norm": 17.015724182128906,
+      "learning_rate": 2.6756756756756757e-06,
+      "loss": 0.3639,
+      "step": 73170
+    },
+    {
+      "epoch": 219.76,
+      "grad_norm": 20.09288215637207,
+      "learning_rate": 2.674674674674675e-06,
+      "loss": 0.3534,
+      "step": 73180
+    },
+    {
+      "epoch": 219.79,
+      "grad_norm": 10.619609832763672,
+      "learning_rate": 2.673673673673674e-06,
+      "loss": 0.3391,
+      "step": 73190
+    },
+    {
+      "epoch": 219.82,
+      "grad_norm": 10.150396347045898,
+      "learning_rate": 2.672672672672673e-06,
+      "loss": 0.329,
+      "step": 73200
+    },
+    {
+      "epoch": 219.85,
+      "grad_norm": 22.929353713989258,
+      "learning_rate": 2.6716716716716717e-06,
+      "loss": 0.3555,
+      "step": 73210
+    },
+    {
+      "epoch": 219.88,
+      "grad_norm": 16.445608139038086,
+      "learning_rate": 2.6706706706706708e-06,
+      "loss": 0.4002,
+      "step": 73220
+    },
+    {
+      "epoch": 219.91,
+      "grad_norm": 12.093541145324707,
+      "learning_rate": 2.66966966966967e-06,
+      "loss": 0.2907,
+      "step": 73230
+    },
+    {
+      "epoch": 219.94,
+      "grad_norm": 14.208827018737793,
+      "learning_rate": 2.6686686686686692e-06,
+      "loss": 0.3695,
+      "step": 73240
+    },
+    {
+      "epoch": 219.97,
+      "grad_norm": 15.175542831420898,
+      "learning_rate": 2.667667667667668e-06,
+      "loss": 0.3844,
+      "step": 73250
+    },
+    {
+      "epoch": 220.0,
+      "grad_norm": 93.89698028564453,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.3703,
+      "step": 73260
+    },
+    {
+      "epoch": 220.0,
+      "eval_accuracy": 0.8641,
+      "eval_loss": 0.547140896320343,
+      "eval_runtime": 12.568,
+      "eval_samples_per_second": 795.674,
+      "eval_steps_per_second": 3.183,
+      "step": 73260
+    },
+    {
+      "epoch": 220.03,
+      "grad_norm": 16.119813919067383,
+      "learning_rate": 2.665665665665666e-06,
+      "loss": 0.3159,
+      "step": 73270
+    },
+    {
+      "epoch": 220.06,
+      "grad_norm": 12.468939781188965,
+      "learning_rate": 2.6646646646646653e-06,
+      "loss": 0.3067,
+      "step": 73280
+    },
+    {
+      "epoch": 220.09,
+      "grad_norm": 12.91759204864502,
+      "learning_rate": 2.663663663663664e-06,
+      "loss": 0.3669,
+      "step": 73290
+    },
+    {
+      "epoch": 220.12,
+      "grad_norm": 15.014217376708984,
+      "learning_rate": 2.662662662662663e-06,
+      "loss": 0.3148,
+      "step": 73300
+    },
+    {
+      "epoch": 220.15,
+      "grad_norm": 14.256143569946289,
+      "learning_rate": 2.661661661661662e-06,
+      "loss": 0.3288,
+      "step": 73310
+    },
+    {
+      "epoch": 220.18,
+      "grad_norm": 13.165700912475586,
+      "learning_rate": 2.660660660660661e-06,
+      "loss": 0.3859,
+      "step": 73320
+    },
+    {
+      "epoch": 220.21,
+      "grad_norm": 22.13396453857422,
+      "learning_rate": 2.6596596596596595e-06,
+      "loss": 0.3219,
+      "step": 73330
+    },
+    {
+      "epoch": 220.24,
+      "grad_norm": 17.448869705200195,
+      "learning_rate": 2.658658658658659e-06,
+      "loss": 0.3412,
+      "step": 73340
+    },
+    {
+      "epoch": 220.27,
+      "grad_norm": 16.61252212524414,
+      "learning_rate": 2.657657657657658e-06,
+      "loss": 0.3347,
+      "step": 73350
+    },
+    {
+      "epoch": 220.3,
+      "grad_norm": 20.672821044921875,
+      "learning_rate": 2.656656656656657e-06,
+      "loss": 0.3186,
+      "step": 73360
+    },
+    {
+      "epoch": 220.33,
+      "grad_norm": 13.950328826904297,
+      "learning_rate": 2.6556556556556556e-06,
+      "loss": 0.3416,
+      "step": 73370
+    },
+    {
+      "epoch": 220.36,
+      "grad_norm": 18.6420955657959,
+      "learning_rate": 2.6546546546546546e-06,
+      "loss": 0.3915,
+      "step": 73380
+    },
+    {
+      "epoch": 220.39,
+      "grad_norm": 15.082834243774414,
+      "learning_rate": 2.653653653653654e-06,
+      "loss": 0.2795,
+      "step": 73390
+    },
+    {
+      "epoch": 220.42,
+      "grad_norm": 18.378875732421875,
+      "learning_rate": 2.652652652652653e-06,
+      "loss": 0.3665,
+      "step": 73400
+    },
+    {
+      "epoch": 220.45,
+      "grad_norm": 17.913249969482422,
+      "learning_rate": 2.6516516516516517e-06,
+      "loss": 0.3585,
+      "step": 73410
+    },
+    {
+      "epoch": 220.48,
+      "grad_norm": 15.53809642791748,
+      "learning_rate": 2.6506506506506507e-06,
+      "loss": 0.336,
+      "step": 73420
+    },
+    {
+      "epoch": 220.51,
+      "grad_norm": 18.03061294555664,
+      "learning_rate": 2.6496496496496497e-06,
+      "loss": 0.3572,
+      "step": 73430
+    },
+    {
+      "epoch": 220.54,
+      "grad_norm": 26.90101432800293,
+      "learning_rate": 2.648648648648649e-06,
+      "loss": 0.3292,
+      "step": 73440
+    },
+    {
+      "epoch": 220.57,
+      "grad_norm": 11.422205924987793,
+      "learning_rate": 2.6476476476476478e-06,
+      "loss": 0.3426,
+      "step": 73450
+    },
+    {
+      "epoch": 220.6,
+      "grad_norm": 15.029540061950684,
+      "learning_rate": 2.6466466466466468e-06,
+      "loss": 0.337,
+      "step": 73460
+    },
+    {
+      "epoch": 220.63,
+      "grad_norm": 23.15026092529297,
+      "learning_rate": 2.645645645645646e-06,
+      "loss": 0.3865,
+      "step": 73470
+    },
+    {
+      "epoch": 220.66,
+      "grad_norm": 21.439027786254883,
+      "learning_rate": 2.644644644644645e-06,
+      "loss": 0.3832,
+      "step": 73480
+    },
+    {
+      "epoch": 220.69,
+      "grad_norm": 12.587804794311523,
+      "learning_rate": 2.6436436436436443e-06,
+      "loss": 0.3772,
+      "step": 73490
+    },
+    {
+      "epoch": 220.72,
+      "grad_norm": 23.89263916015625,
+      "learning_rate": 2.642642642642643e-06,
+      "loss": 0.329,
+      "step": 73500
+    },
+    {
+      "epoch": 220.75,
+      "grad_norm": 14.361783981323242,
+      "learning_rate": 2.641641641641642e-06,
+      "loss": 0.364,
+      "step": 73510
+    },
+    {
+      "epoch": 220.78,
+      "grad_norm": 10.89684772491455,
+      "learning_rate": 2.640640640640641e-06,
+      "loss": 0.3473,
+      "step": 73520
+    },
+    {
+      "epoch": 220.81,
+      "grad_norm": 16.636632919311523,
+      "learning_rate": 2.63963963963964e-06,
+      "loss": 0.3654,
+      "step": 73530
+    },
+    {
+      "epoch": 220.84,
+      "grad_norm": 21.115739822387695,
+      "learning_rate": 2.6386386386386385e-06,
+      "loss": 0.3871,
+      "step": 73540
+    },
+    {
+      "epoch": 220.87,
+      "grad_norm": 19.39952850341797,
+      "learning_rate": 2.637637637637638e-06,
+      "loss": 0.339,
+      "step": 73550
+    },
+    {
+      "epoch": 220.9,
+      "grad_norm": 16.43108367919922,
+      "learning_rate": 2.636636636636637e-06,
+      "loss": 0.341,
+      "step": 73560
+    },
+    {
+      "epoch": 220.93,
+      "grad_norm": 12.98720932006836,
+      "learning_rate": 2.635635635635636e-06,
+      "loss": 0.3158,
+      "step": 73570
+    },
+    {
+      "epoch": 220.96,
+      "grad_norm": 27.039121627807617,
+      "learning_rate": 2.6346346346346346e-06,
+      "loss": 0.351,
+      "step": 73580
+    },
+    {
+      "epoch": 220.99,
+      "grad_norm": 26.872028350830078,
+      "learning_rate": 2.6336336336336336e-06,
+      "loss": 0.3318,
+      "step": 73590
+    },
+    {
+      "epoch": 221.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5431638956069946,
+      "eval_runtime": 12.5344,
+      "eval_samples_per_second": 797.804,
+      "eval_steps_per_second": 3.191,
+      "step": 73593
+    },
+    {
+      "epoch": 221.02,
+      "grad_norm": 7.857973098754883,
+      "learning_rate": 2.632632632632633e-06,
+      "loss": 0.2851,
+      "step": 73600
+    },
+    {
+      "epoch": 221.05,
+      "grad_norm": 16.360774993896484,
+      "learning_rate": 2.631631631631632e-06,
+      "loss": 0.3841,
+      "step": 73610
+    },
+    {
+      "epoch": 221.08,
+      "grad_norm": 14.83067798614502,
+      "learning_rate": 2.6306306306306306e-06,
+      "loss": 0.3451,
+      "step": 73620
+    },
+    {
+      "epoch": 221.11,
+      "grad_norm": 12.677286148071289,
+      "learning_rate": 2.6296296296296297e-06,
+      "loss": 0.3417,
+      "step": 73630
+    },
+    {
+      "epoch": 221.14,
+      "grad_norm": 17.7976131439209,
+      "learning_rate": 2.6286286286286287e-06,
+      "loss": 0.3694,
+      "step": 73640
+    },
+    {
+      "epoch": 221.17,
+      "grad_norm": 13.760993957519531,
+      "learning_rate": 2.627627627627628e-06,
+      "loss": 0.338,
+      "step": 73650
+    },
+    {
+      "epoch": 221.2,
+      "grad_norm": 14.099915504455566,
+      "learning_rate": 2.6266266266266267e-06,
+      "loss": 0.298,
+      "step": 73660
+    },
+    {
+      "epoch": 221.23,
+      "grad_norm": 17.142290115356445,
+      "learning_rate": 2.6256256256256257e-06,
+      "loss": 0.3261,
+      "step": 73670
+    },
+    {
+      "epoch": 221.26,
+      "grad_norm": 11.732195854187012,
+      "learning_rate": 2.6246246246246248e-06,
+      "loss": 0.2858,
+      "step": 73680
+    },
+    {
+      "epoch": 221.29,
+      "grad_norm": 18.57823944091797,
+      "learning_rate": 2.623623623623624e-06,
+      "loss": 0.3642,
+      "step": 73690
+    },
+    {
+      "epoch": 221.32,
+      "grad_norm": 18.096223831176758,
+      "learning_rate": 2.6226226226226224e-06,
+      "loss": 0.3271,
+      "step": 73700
+    },
+    {
+      "epoch": 221.35,
+      "grad_norm": 19.478445053100586,
+      "learning_rate": 2.621621621621622e-06,
+      "loss": 0.3566,
+      "step": 73710
+    },
+    {
+      "epoch": 221.38,
+      "grad_norm": 16.64015007019043,
+      "learning_rate": 2.620620620620621e-06,
+      "loss": 0.3316,
+      "step": 73720
+    },
+    {
+      "epoch": 221.41,
+      "grad_norm": 19.012916564941406,
+      "learning_rate": 2.61961961961962e-06,
+      "loss": 0.3491,
+      "step": 73730
+    },
+    {
+      "epoch": 221.44,
+      "grad_norm": 14.754829406738281,
+      "learning_rate": 2.6186186186186193e-06,
+      "loss": 0.3368,
+      "step": 73740
+    },
+    {
+      "epoch": 221.47,
+      "grad_norm": 16.7619686126709,
+      "learning_rate": 2.617617617617618e-06,
+      "loss": 0.3549,
+      "step": 73750
+    },
+    {
+      "epoch": 221.5,
+      "grad_norm": 21.964725494384766,
+      "learning_rate": 2.616616616616617e-06,
+      "loss": 0.4015,
+      "step": 73760
+    },
+    {
+      "epoch": 221.53,
+      "grad_norm": 13.253177642822266,
+      "learning_rate": 2.615615615615616e-06,
+      "loss": 0.3021,
+      "step": 73770
+    },
+    {
+      "epoch": 221.56,
+      "grad_norm": 14.414772033691406,
+      "learning_rate": 2.614614614614615e-06,
+      "loss": 0.3092,
+      "step": 73780
+    },
+    {
+      "epoch": 221.59,
+      "grad_norm": 21.940547943115234,
+      "learning_rate": 2.6136136136136135e-06,
+      "loss": 0.3495,
+      "step": 73790
+    },
+    {
+      "epoch": 221.62,
+      "grad_norm": 17.241363525390625,
+      "learning_rate": 2.612612612612613e-06,
+      "loss": 0.4025,
+      "step": 73800
+    },
+    {
+      "epoch": 221.65,
+      "grad_norm": 13.235084533691406,
+      "learning_rate": 2.611611611611612e-06,
+      "loss": 0.36,
+      "step": 73810
+    },
+    {
+      "epoch": 221.68,
+      "grad_norm": 17.542455673217773,
+      "learning_rate": 2.610610610610611e-06,
+      "loss": 0.342,
+      "step": 73820
+    },
+    {
+      "epoch": 221.71,
+      "grad_norm": 18.5478515625,
+      "learning_rate": 2.6096096096096096e-06,
+      "loss": 0.3627,
+      "step": 73830
+    },
+    {
+      "epoch": 221.74,
+      "grad_norm": 25.21489715576172,
+      "learning_rate": 2.6086086086086086e-06,
+      "loss": 0.3799,
+      "step": 73840
+    },
+    {
+      "epoch": 221.77,
+      "grad_norm": 11.910326957702637,
+      "learning_rate": 2.607607607607608e-06,
+      "loss": 0.3376,
+      "step": 73850
+    },
+    {
+      "epoch": 221.8,
+      "grad_norm": 16.475461959838867,
+      "learning_rate": 2.606606606606607e-06,
+      "loss": 0.3497,
+      "step": 73860
+    },
+    {
+      "epoch": 221.83,
+      "grad_norm": 17.324541091918945,
+      "learning_rate": 2.6056056056056057e-06,
+      "loss": 0.3489,
+      "step": 73870
+    },
+    {
+      "epoch": 221.86,
+      "grad_norm": 21.99091339111328,
+      "learning_rate": 2.6046046046046047e-06,
+      "loss": 0.3667,
+      "step": 73880
+    },
+    {
+      "epoch": 221.89,
+      "grad_norm": 13.620786666870117,
+      "learning_rate": 2.6036036036036037e-06,
+      "loss": 0.3819,
+      "step": 73890
+    },
+    {
+      "epoch": 221.92,
+      "grad_norm": 14.649778366088867,
+      "learning_rate": 2.602602602602603e-06,
+      "loss": 0.3516,
+      "step": 73900
+    },
+    {
+      "epoch": 221.95,
+      "grad_norm": 15.421683311462402,
+      "learning_rate": 2.6016016016016017e-06,
+      "loss": 0.3493,
+      "step": 73910
+    },
+    {
+      "epoch": 221.98,
+      "grad_norm": 12.94367504119873,
+      "learning_rate": 2.6006006006006008e-06,
+      "loss": 0.3573,
+      "step": 73920
+    },
+    {
+      "epoch": 222.0,
+      "eval_accuracy": 0.8631,
+      "eval_loss": 0.5473458170890808,
+      "eval_runtime": 13.1186,
+      "eval_samples_per_second": 762.276,
+      "eval_steps_per_second": 3.049,
+      "step": 73926
+    },
+    {
+      "epoch": 222.01,
+      "grad_norm": 12.056336402893066,
+      "learning_rate": 2.5995995995995998e-06,
+      "loss": 0.4102,
+      "step": 73930
+    },
+    {
+      "epoch": 222.04,
+      "grad_norm": 23.42940330505371,
+      "learning_rate": 2.598598598598599e-06,
+      "loss": 0.3602,
+      "step": 73940
+    },
+    {
+      "epoch": 222.07,
+      "grad_norm": 11.72469711303711,
+      "learning_rate": 2.5975975975975974e-06,
+      "loss": 0.3293,
+      "step": 73950
+    },
+    {
+      "epoch": 222.1,
+      "grad_norm": 12.916447639465332,
+      "learning_rate": 2.596596596596597e-06,
+      "loss": 0.358,
+      "step": 73960
+    },
+    {
+      "epoch": 222.13,
+      "grad_norm": 12.716425895690918,
+      "learning_rate": 2.595595595595596e-06,
+      "loss": 0.2769,
+      "step": 73970
+    },
+    {
+      "epoch": 222.16,
+      "grad_norm": 12.093764305114746,
+      "learning_rate": 2.594594594594595e-06,
+      "loss": 0.3522,
+      "step": 73980
+    },
+    {
+      "epoch": 222.19,
+      "grad_norm": 16.70994758605957,
+      "learning_rate": 2.5935935935935935e-06,
+      "loss": 0.3162,
+      "step": 73990
+    },
+    {
+      "epoch": 222.22,
+      "grad_norm": 13.669795989990234,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.3523,
+      "step": 74000
+    },
+    {
+      "epoch": 222.25,
+      "grad_norm": 15.199612617492676,
+      "learning_rate": 2.591591591591592e-06,
+      "loss": 0.3254,
+      "step": 74010
+    },
+    {
+      "epoch": 222.28,
+      "grad_norm": 16.216768264770508,
+      "learning_rate": 2.590590590590591e-06,
+      "loss": 0.4133,
+      "step": 74020
+    },
+    {
+      "epoch": 222.31,
+      "grad_norm": 21.008953094482422,
+      "learning_rate": 2.58958958958959e-06,
+      "loss": 0.3833,
+      "step": 74030
+    },
+    {
+      "epoch": 222.34,
+      "grad_norm": 16.82625961303711,
+      "learning_rate": 2.5885885885885886e-06,
+      "loss": 0.3419,
+      "step": 74040
+    },
+    {
+      "epoch": 222.37,
+      "grad_norm": 17.64153289794922,
+      "learning_rate": 2.5875875875875876e-06,
+      "loss": 0.336,
+      "step": 74050
+    },
+    {
+      "epoch": 222.4,
+      "grad_norm": 13.105913162231445,
+      "learning_rate": 2.586586586586587e-06,
+      "loss": 0.2861,
+      "step": 74060
+    },
+    {
+      "epoch": 222.43,
+      "grad_norm": 15.945516586303711,
+      "learning_rate": 2.585585585585586e-06,
+      "loss": 0.3878,
+      "step": 74070
+    },
+    {
+      "epoch": 222.46,
+      "grad_norm": 12.69954776763916,
+      "learning_rate": 2.5845845845845846e-06,
+      "loss": 0.3614,
+      "step": 74080
+    },
+    {
+      "epoch": 222.49,
+      "grad_norm": 12.148127555847168,
+      "learning_rate": 2.5835835835835837e-06,
+      "loss": 0.3326,
+      "step": 74090
+    },
+    {
+      "epoch": 222.52,
+      "grad_norm": 20.88222885131836,
+      "learning_rate": 2.5825825825825827e-06,
+      "loss": 0.3598,
+      "step": 74100
+    },
+    {
+      "epoch": 222.55,
+      "grad_norm": 14.098403930664062,
+      "learning_rate": 2.581581581581582e-06,
+      "loss": 0.3402,
+      "step": 74110
+    },
+    {
+      "epoch": 222.58,
+      "grad_norm": 17.07585906982422,
+      "learning_rate": 2.5805805805805807e-06,
+      "loss": 0.3466,
+      "step": 74120
+    },
+    {
+      "epoch": 222.61,
+      "grad_norm": 25.63220977783203,
+      "learning_rate": 2.5795795795795797e-06,
+      "loss": 0.4028,
+      "step": 74130
+    },
+    {
+      "epoch": 222.64,
+      "grad_norm": 13.469189643859863,
+      "learning_rate": 2.5785785785785787e-06,
+      "loss": 0.3388,
+      "step": 74140
+    },
+    {
+      "epoch": 222.67,
+      "grad_norm": 29.968055725097656,
+      "learning_rate": 2.577577577577578e-06,
+      "loss": 0.3545,
+      "step": 74150
+    },
+    {
+      "epoch": 222.7,
+      "grad_norm": 11.103266716003418,
+      "learning_rate": 2.5765765765765764e-06,
+      "loss": 0.3851,
+      "step": 74160
+    },
+    {
+      "epoch": 222.73,
+      "grad_norm": 13.595907211303711,
+      "learning_rate": 2.575575575575576e-06,
+      "loss": 0.3489,
+      "step": 74170
+    },
+    {
+      "epoch": 222.76,
+      "grad_norm": 12.351510047912598,
+      "learning_rate": 2.574574574574575e-06,
+      "loss": 0.3136,
+      "step": 74180
+    },
+    {
+      "epoch": 222.79,
+      "grad_norm": 19.371864318847656,
+      "learning_rate": 2.573573573573574e-06,
+      "loss": 0.3619,
+      "step": 74190
+    },
+    {
+      "epoch": 222.82,
+      "grad_norm": 13.918539047241211,
+      "learning_rate": 2.5725725725725724e-06,
+      "loss": 0.3324,
+      "step": 74200
+    },
+    {
+      "epoch": 222.85,
+      "grad_norm": 15.874768257141113,
+      "learning_rate": 2.571571571571572e-06,
+      "loss": 0.3643,
+      "step": 74210
+    },
+    {
+      "epoch": 222.88,
+      "grad_norm": 18.463224411010742,
+      "learning_rate": 2.570570570570571e-06,
+      "loss": 0.3381,
+      "step": 74220
+    },
+    {
+      "epoch": 222.91,
+      "grad_norm": 18.725080490112305,
+      "learning_rate": 2.56956956956957e-06,
+      "loss": 0.4022,
+      "step": 74230
+    },
+    {
+      "epoch": 222.94,
+      "grad_norm": 13.277693748474121,
+      "learning_rate": 2.5685685685685685e-06,
+      "loss": 0.3677,
+      "step": 74240
+    },
+    {
+      "epoch": 222.97,
+      "grad_norm": 16.976743698120117,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 0.3308,
+      "step": 74250
+    },
+    {
+      "epoch": 223.0,
+      "eval_accuracy": 0.8663,
+      "eval_loss": 0.544793426990509,
+      "eval_runtime": 12.7349,
+      "eval_samples_per_second": 785.242,
+      "eval_steps_per_second": 3.141,
+      "step": 74259
+    },
+    {
+      "epoch": 223.0,
+      "grad_norm": 12.211898803710938,
+      "learning_rate": 2.566566566566567e-06,
+      "loss": 0.3534,
+      "step": 74260
+    },
+    {
+      "epoch": 223.03,
+      "grad_norm": 22.09015464782715,
+      "learning_rate": 2.565565565565566e-06,
+      "loss": 0.2935,
+      "step": 74270
+    },
+    {
+      "epoch": 223.06,
+      "grad_norm": 15.301719665527344,
+      "learning_rate": 2.564564564564565e-06,
+      "loss": 0.3056,
+      "step": 74280
+    },
+    {
+      "epoch": 223.09,
+      "grad_norm": 15.730999946594238,
+      "learning_rate": 2.5635635635635636e-06,
+      "loss": 0.3477,
+      "step": 74290
+    },
+    {
+      "epoch": 223.12,
+      "grad_norm": 17.41704559326172,
+      "learning_rate": 2.5625625625625626e-06,
+      "loss": 0.3249,
+      "step": 74300
+    },
+    {
+      "epoch": 223.15,
+      "grad_norm": 14.325186729431152,
+      "learning_rate": 2.561561561561562e-06,
+      "loss": 0.3023,
+      "step": 74310
+    },
+    {
+      "epoch": 223.18,
+      "grad_norm": 21.246700286865234,
+      "learning_rate": 2.560560560560561e-06,
+      "loss": 0.3084,
+      "step": 74320
+    },
+    {
+      "epoch": 223.21,
+      "grad_norm": 22.112545013427734,
+      "learning_rate": 2.5595595595595597e-06,
+      "loss": 0.3778,
+      "step": 74330
+    },
+    {
+      "epoch": 223.24,
+      "grad_norm": 20.665559768676758,
+      "learning_rate": 2.5585585585585587e-06,
+      "loss": 0.3253,
+      "step": 74340
+    },
+    {
+      "epoch": 223.27,
+      "grad_norm": 12.82650375366211,
+      "learning_rate": 2.5575575575575577e-06,
+      "loss": 0.3814,
+      "step": 74350
+    },
+    {
+      "epoch": 223.3,
+      "grad_norm": 13.541102409362793,
+      "learning_rate": 2.556556556556557e-06,
+      "loss": 0.3115,
+      "step": 74360
+    },
+    {
+      "epoch": 223.33,
+      "grad_norm": 13.991399765014648,
+      "learning_rate": 2.5555555555555557e-06,
+      "loss": 0.325,
+      "step": 74370
+    },
+    {
+      "epoch": 223.36,
+      "grad_norm": 19.77086067199707,
+      "learning_rate": 2.5545545545545548e-06,
+      "loss": 0.3387,
+      "step": 74380
+    },
+    {
+      "epoch": 223.39,
+      "grad_norm": 15.434678077697754,
+      "learning_rate": 2.5535535535535538e-06,
+      "loss": 0.3678,
+      "step": 74390
+    },
+    {
+      "epoch": 223.42,
+      "grad_norm": 12.814022064208984,
+      "learning_rate": 2.552552552552553e-06,
+      "loss": 0.2978,
+      "step": 74400
+    },
+    {
+      "epoch": 223.45,
+      "grad_norm": 19.63626480102539,
+      "learning_rate": 2.5515515515515514e-06,
+      "loss": 0.3401,
+      "step": 74410
+    },
+    {
+      "epoch": 223.48,
+      "grad_norm": 13.059757232666016,
+      "learning_rate": 2.550550550550551e-06,
+      "loss": 0.3068,
+      "step": 74420
+    },
+    {
+      "epoch": 223.51,
+      "grad_norm": 15.182046890258789,
+      "learning_rate": 2.54954954954955e-06,
+      "loss": 0.3964,
+      "step": 74430
+    },
+    {
+      "epoch": 223.54,
+      "grad_norm": 12.037517547607422,
+      "learning_rate": 2.548548548548549e-06,
+      "loss": 0.3319,
+      "step": 74440
+    },
+    {
+      "epoch": 223.57,
+      "grad_norm": 14.350536346435547,
+      "learning_rate": 2.5475475475475475e-06,
+      "loss": 0.3397,
+      "step": 74450
+    },
+    {
+      "epoch": 223.6,
+      "grad_norm": 23.33458137512207,
+      "learning_rate": 2.5465465465465465e-06,
+      "loss": 0.3491,
+      "step": 74460
+    },
+    {
+      "epoch": 223.63,
+      "grad_norm": 15.749000549316406,
+      "learning_rate": 2.545545545545546e-06,
+      "loss": 0.362,
+      "step": 74470
+    },
+    {
+      "epoch": 223.66,
+      "grad_norm": 19.656980514526367,
+      "learning_rate": 2.544544544544545e-06,
+      "loss": 0.3779,
+      "step": 74480
+    },
+    {
+      "epoch": 223.69,
+      "grad_norm": 16.000106811523438,
+      "learning_rate": 2.5435435435435435e-06,
+      "loss": 0.3701,
+      "step": 74490
+    },
+    {
+      "epoch": 223.72,
+      "grad_norm": 13.005067825317383,
+      "learning_rate": 2.5425425425425426e-06,
+      "loss": 0.3235,
+      "step": 74500
+    },
+    {
+      "epoch": 223.75,
+      "grad_norm": 23.284536361694336,
+      "learning_rate": 2.5415415415415416e-06,
+      "loss": 0.3744,
+      "step": 74510
+    },
+    {
+      "epoch": 223.78,
+      "grad_norm": 13.183226585388184,
+      "learning_rate": 2.540540540540541e-06,
+      "loss": 0.3615,
+      "step": 74520
+    },
+    {
+      "epoch": 223.81,
+      "grad_norm": 22.76864242553711,
+      "learning_rate": 2.53953953953954e-06,
+      "loss": 0.3675,
+      "step": 74530
+    },
+    {
+      "epoch": 223.84,
+      "grad_norm": 18.684398651123047,
+      "learning_rate": 2.5385385385385386e-06,
+      "loss": 0.3546,
+      "step": 74540
+    },
+    {
+      "epoch": 223.87,
+      "grad_norm": 11.809819221496582,
+      "learning_rate": 2.5375375375375377e-06,
+      "loss": 0.3341,
+      "step": 74550
+    },
+    {
+      "epoch": 223.9,
+      "grad_norm": 20.473600387573242,
+      "learning_rate": 2.5365365365365367e-06,
+      "loss": 0.356,
+      "step": 74560
+    },
+    {
+      "epoch": 223.93,
+      "grad_norm": 14.810294151306152,
+      "learning_rate": 2.535535535535536e-06,
+      "loss": 0.3241,
+      "step": 74570
+    },
+    {
+      "epoch": 223.96,
+      "grad_norm": 18.167728424072266,
+      "learning_rate": 2.5345345345345347e-06,
+      "loss": 0.3461,
+      "step": 74580
+    },
+    {
+      "epoch": 223.99,
+      "grad_norm": 13.545295715332031,
+      "learning_rate": 2.5335335335335337e-06,
+      "loss": 0.3329,
+      "step": 74590
+    },
+    {
+      "epoch": 224.0,
+      "eval_accuracy": 0.8635,
+      "eval_loss": 0.5444658398628235,
+      "eval_runtime": 12.9867,
+      "eval_samples_per_second": 770.019,
+      "eval_steps_per_second": 3.08,
+      "step": 74592
+    },
+    {
+      "epoch": 224.02,
+      "grad_norm": 9.197989463806152,
+      "learning_rate": 2.5325325325325327e-06,
+      "loss": 0.4172,
+      "step": 74600
+    },
+    {
+      "epoch": 224.05,
+      "grad_norm": 15.712944030761719,
+      "learning_rate": 2.5315315315315318e-06,
+      "loss": 0.4015,
+      "step": 74610
+    },
+    {
+      "epoch": 224.08,
+      "grad_norm": 14.936673164367676,
+      "learning_rate": 2.5305305305305304e-06,
+      "loss": 0.2997,
+      "step": 74620
+    },
+    {
+      "epoch": 224.11,
+      "grad_norm": 13.022282600402832,
+      "learning_rate": 2.52952952952953e-06,
+      "loss": 0.3366,
+      "step": 74630
+    },
+    {
+      "epoch": 224.14,
+      "grad_norm": 12.810697555541992,
+      "learning_rate": 2.528528528528529e-06,
+      "loss": 0.3169,
+      "step": 74640
+    },
+    {
+      "epoch": 224.17,
+      "grad_norm": 15.988986015319824,
+      "learning_rate": 2.527527527527528e-06,
+      "loss": 0.3606,
+      "step": 74650
+    },
+    {
+      "epoch": 224.2,
+      "grad_norm": 22.866680145263672,
+      "learning_rate": 2.5265265265265264e-06,
+      "loss": 0.3694,
+      "step": 74660
+    },
+    {
+      "epoch": 224.23,
+      "grad_norm": 17.344297409057617,
+      "learning_rate": 2.5255255255255254e-06,
+      "loss": 0.3439,
+      "step": 74670
+    },
+    {
+      "epoch": 224.26,
+      "grad_norm": 18.194665908813477,
+      "learning_rate": 2.524524524524525e-06,
+      "loss": 0.3129,
+      "step": 74680
+    },
+    {
+      "epoch": 224.29,
+      "grad_norm": 13.91851806640625,
+      "learning_rate": 2.523523523523524e-06,
+      "loss": 0.3927,
+      "step": 74690
+    },
+    {
+      "epoch": 224.32,
+      "grad_norm": 10.205693244934082,
+      "learning_rate": 2.5225225225225225e-06,
+      "loss": 0.321,
+      "step": 74700
+    },
+    {
+      "epoch": 224.35,
+      "grad_norm": 17.198246002197266,
+      "learning_rate": 2.5215215215215215e-06,
+      "loss": 0.3213,
+      "step": 74710
+    },
+    {
+      "epoch": 224.38,
+      "grad_norm": 20.905187606811523,
+      "learning_rate": 2.520520520520521e-06,
+      "loss": 0.3365,
+      "step": 74720
+    },
+    {
+      "epoch": 224.41,
+      "grad_norm": 11.85018253326416,
+      "learning_rate": 2.51951951951952e-06,
+      "loss": 0.3771,
+      "step": 74730
+    },
+    {
+      "epoch": 224.44,
+      "grad_norm": 13.839212417602539,
+      "learning_rate": 2.5185185185185186e-06,
+      "loss": 0.3293,
+      "step": 74740
+    },
+    {
+      "epoch": 224.47,
+      "grad_norm": 11.289928436279297,
+      "learning_rate": 2.5175175175175176e-06,
+      "loss": 0.3056,
+      "step": 74750
+    },
+    {
+      "epoch": 224.5,
+      "grad_norm": 14.45038890838623,
+      "learning_rate": 2.5165165165165166e-06,
+      "loss": 0.3455,
+      "step": 74760
+    },
+    {
+      "epoch": 224.53,
+      "grad_norm": 16.99696159362793,
+      "learning_rate": 2.515515515515516e-06,
+      "loss": 0.3365,
+      "step": 74770
+    },
+    {
+      "epoch": 224.56,
+      "grad_norm": 13.112373352050781,
+      "learning_rate": 2.5145145145145147e-06,
+      "loss": 0.3253,
+      "step": 74780
+    },
+    {
+      "epoch": 224.59,
+      "grad_norm": 16.063587188720703,
+      "learning_rate": 2.5135135135135137e-06,
+      "loss": 0.3668,
+      "step": 74790
+    },
+    {
+      "epoch": 224.62,
+      "grad_norm": 23.118633270263672,
+      "learning_rate": 2.5125125125125127e-06,
+      "loss": 0.3085,
+      "step": 74800
+    },
+    {
+      "epoch": 224.65,
+      "grad_norm": 10.37130069732666,
+      "learning_rate": 2.5115115115115117e-06,
+      "loss": 0.311,
+      "step": 74810
+    },
+    {
+      "epoch": 224.68,
+      "grad_norm": 28.902193069458008,
+      "learning_rate": 2.510510510510511e-06,
+      "loss": 0.3641,
+      "step": 74820
+    },
+    {
+      "epoch": 224.71,
+      "grad_norm": 29.54743003845215,
+      "learning_rate": 2.5095095095095097e-06,
+      "loss": 0.2963,
+      "step": 74830
+    },
+    {
+      "epoch": 224.74,
+      "grad_norm": 17.26615333557129,
+      "learning_rate": 2.5085085085085088e-06,
+      "loss": 0.3983,
+      "step": 74840
+    },
+    {
+      "epoch": 224.77,
+      "grad_norm": 23.9836483001709,
+      "learning_rate": 2.5075075075075078e-06,
+      "loss": 0.3507,
+      "step": 74850
+    },
+    {
+      "epoch": 224.8,
+      "grad_norm": 21.089109420776367,
+      "learning_rate": 2.506506506506507e-06,
+      "loss": 0.3498,
+      "step": 74860
+    },
+    {
+      "epoch": 224.83,
+      "grad_norm": 15.913220405578613,
+      "learning_rate": 2.5055055055055054e-06,
+      "loss": 0.2787,
+      "step": 74870
+    },
+    {
+      "epoch": 224.86,
+      "grad_norm": 12.83398151397705,
+      "learning_rate": 2.504504504504505e-06,
+      "loss": 0.3236,
+      "step": 74880
+    },
+    {
+      "epoch": 224.89,
+      "grad_norm": 11.772079467773438,
+      "learning_rate": 2.503503503503504e-06,
+      "loss": 0.317,
+      "step": 74890
+    },
+    {
+      "epoch": 224.92,
+      "grad_norm": 13.901806831359863,
+      "learning_rate": 2.502502502502503e-06,
+      "loss": 0.3506,
+      "step": 74900
+    },
+    {
+      "epoch": 224.95,
+      "grad_norm": 18.30997657775879,
+      "learning_rate": 2.5015015015015015e-06,
+      "loss": 0.3612,
+      "step": 74910
+    },
+    {
+      "epoch": 224.98,
+      "grad_norm": 11.142070770263672,
+      "learning_rate": 2.5005005005005005e-06,
+      "loss": 0.3429,
+      "step": 74920
+    },
+    {
+      "epoch": 225.0,
+      "eval_accuracy": 0.8631,
+      "eval_loss": 0.5444555282592773,
+      "eval_runtime": 12.9027,
+      "eval_samples_per_second": 775.03,
+      "eval_steps_per_second": 3.1,
+      "step": 74925
+    },
+    {
+      "epoch": 225.02,
+      "grad_norm": 10.432902336120605,
+      "learning_rate": 2.4994994994995e-06,
+      "loss": 0.3408,
+      "step": 74930
+    },
+    {
+      "epoch": 225.05,
+      "grad_norm": 17.60347557067871,
+      "learning_rate": 2.4984984984984985e-06,
+      "loss": 0.3278,
+      "step": 74940
+    },
+    {
+      "epoch": 225.08,
+      "grad_norm": 11.254010200500488,
+      "learning_rate": 2.497497497497498e-06,
+      "loss": 0.339,
+      "step": 74950
+    },
+    {
+      "epoch": 225.11,
+      "grad_norm": 12.723819732666016,
+      "learning_rate": 2.4964964964964966e-06,
+      "loss": 0.3376,
+      "step": 74960
+    },
+    {
+      "epoch": 225.14,
+      "grad_norm": 13.788206100463867,
+      "learning_rate": 2.4954954954954956e-06,
+      "loss": 0.3438,
+      "step": 74970
+    },
+    {
+      "epoch": 225.17,
+      "grad_norm": 10.995004653930664,
+      "learning_rate": 2.4944944944944946e-06,
+      "loss": 0.3359,
+      "step": 74980
+    },
+    {
+      "epoch": 225.2,
+      "grad_norm": 16.334293365478516,
+      "learning_rate": 2.4934934934934936e-06,
+      "loss": 0.3525,
+      "step": 74990
+    },
+    {
+      "epoch": 225.23,
+      "grad_norm": 15.86385726928711,
+      "learning_rate": 2.4924924924924926e-06,
+      "loss": 0.3652,
+      "step": 75000
+    },
+    {
+      "epoch": 225.26,
+      "grad_norm": 15.741164207458496,
+      "learning_rate": 2.4914914914914916e-06,
+      "loss": 0.3696,
+      "step": 75010
+    },
+    {
+      "epoch": 225.29,
+      "grad_norm": 11.4758939743042,
+      "learning_rate": 2.4904904904904907e-06,
+      "loss": 0.2861,
+      "step": 75020
+    },
+    {
+      "epoch": 225.32,
+      "grad_norm": 13.456703186035156,
+      "learning_rate": 2.4894894894894897e-06,
+      "loss": 0.3541,
+      "step": 75030
+    },
+    {
+      "epoch": 225.35,
+      "grad_norm": 15.073731422424316,
+      "learning_rate": 2.4884884884884887e-06,
+      "loss": 0.3574,
+      "step": 75040
+    },
+    {
+      "epoch": 225.38,
+      "grad_norm": 14.87912654876709,
+      "learning_rate": 2.4874874874874877e-06,
+      "loss": 0.3003,
+      "step": 75050
+    },
+    {
+      "epoch": 225.41,
+      "grad_norm": 20.75531768798828,
+      "learning_rate": 2.4864864864864867e-06,
+      "loss": 0.3459,
+      "step": 75060
+    },
+    {
+      "epoch": 225.44,
+      "grad_norm": 14.375066757202148,
+      "learning_rate": 2.4854854854854858e-06,
+      "loss": 0.2987,
+      "step": 75070
+    },
+    {
+      "epoch": 225.47,
+      "grad_norm": 22.751419067382812,
+      "learning_rate": 2.4844844844844848e-06,
+      "loss": 0.3743,
+      "step": 75080
+    },
+    {
+      "epoch": 225.5,
+      "grad_norm": 19.741411209106445,
+      "learning_rate": 2.483483483483484e-06,
+      "loss": 0.3258,
+      "step": 75090
+    },
+    {
+      "epoch": 225.53,
+      "grad_norm": 15.151894569396973,
+      "learning_rate": 2.482482482482483e-06,
+      "loss": 0.3148,
+      "step": 75100
+    },
+    {
+      "epoch": 225.56,
+      "grad_norm": 20.170822143554688,
+      "learning_rate": 2.481481481481482e-06,
+      "loss": 0.3464,
+      "step": 75110
+    },
+    {
+      "epoch": 225.59,
+      "grad_norm": 25.123380661010742,
+      "learning_rate": 2.480480480480481e-06,
+      "loss": 0.3716,
+      "step": 75120
+    },
+    {
+      "epoch": 225.62,
+      "grad_norm": 17.66338539123535,
+      "learning_rate": 2.4794794794794794e-06,
+      "loss": 0.3271,
+      "step": 75130
+    },
+    {
+      "epoch": 225.65,
+      "grad_norm": 30.495222091674805,
+      "learning_rate": 2.478478478478479e-06,
+      "loss": 0.3209,
+      "step": 75140
+    },
+    {
+      "epoch": 225.68,
+      "grad_norm": 10.102629661560059,
+      "learning_rate": 2.4774774774774775e-06,
+      "loss": 0.35,
+      "step": 75150
+    },
+    {
+      "epoch": 225.71,
+      "grad_norm": 21.29137420654297,
+      "learning_rate": 2.476476476476477e-06,
+      "loss": 0.3643,
+      "step": 75160
+    },
+    {
+      "epoch": 225.74,
+      "grad_norm": 13.720152854919434,
+      "learning_rate": 2.4754754754754755e-06,
+      "loss": 0.3529,
+      "step": 75170
+    },
+    {
+      "epoch": 225.77,
+      "grad_norm": 13.489992141723633,
+      "learning_rate": 2.474474474474475e-06,
+      "loss": 0.3438,
+      "step": 75180
+    },
+    {
+      "epoch": 225.8,
+      "grad_norm": 14.043858528137207,
+      "learning_rate": 2.4734734734734736e-06,
+      "loss": 0.3473,
+      "step": 75190
+    },
+    {
+      "epoch": 225.83,
+      "grad_norm": 10.892147064208984,
+      "learning_rate": 2.4724724724724726e-06,
+      "loss": 0.3428,
+      "step": 75200
+    },
+    {
+      "epoch": 225.86,
+      "grad_norm": 26.266714096069336,
+      "learning_rate": 2.4714714714714716e-06,
+      "loss": 0.3181,
+      "step": 75210
+    },
+    {
+      "epoch": 225.89,
+      "grad_norm": 22.564538955688477,
+      "learning_rate": 2.4704704704704706e-06,
+      "loss": 0.2984,
+      "step": 75220
+    },
+    {
+      "epoch": 225.92,
+      "grad_norm": 14.462691307067871,
+      "learning_rate": 2.4694694694694696e-06,
+      "loss": 0.3135,
+      "step": 75230
+    },
+    {
+      "epoch": 225.95,
+      "grad_norm": 13.653186798095703,
+      "learning_rate": 2.4684684684684686e-06,
+      "loss": 0.3628,
+      "step": 75240
+    },
+    {
+      "epoch": 225.98,
+      "grad_norm": 21.204233169555664,
+      "learning_rate": 2.4674674674674677e-06,
+      "loss": 0.3494,
+      "step": 75250
+    },
+    {
+      "epoch": 226.0,
+      "eval_accuracy": 0.8632,
+      "eval_loss": 0.5433376431465149,
+      "eval_runtime": 12.7343,
+      "eval_samples_per_second": 785.281,
+      "eval_steps_per_second": 3.141,
+      "step": 75258
+    },
+    {
+      "epoch": 226.01,
+      "grad_norm": 14.082113265991211,
+      "learning_rate": 2.4664664664664667e-06,
+      "loss": 0.3438,
+      "step": 75260
+    },
+    {
+      "epoch": 226.04,
+      "grad_norm": 21.770261764526367,
+      "learning_rate": 2.4654654654654657e-06,
+      "loss": 0.327,
+      "step": 75270
+    },
+    {
+      "epoch": 226.07,
+      "grad_norm": 11.517925262451172,
+      "learning_rate": 2.4644644644644647e-06,
+      "loss": 0.3528,
+      "step": 75280
+    },
+    {
+      "epoch": 226.1,
+      "grad_norm": 13.681634902954102,
+      "learning_rate": 2.4634634634634637e-06,
+      "loss": 0.3354,
+      "step": 75290
+    },
+    {
+      "epoch": 226.13,
+      "grad_norm": 15.576708793640137,
+      "learning_rate": 2.4624624624624628e-06,
+      "loss": 0.3504,
+      "step": 75300
+    },
+    {
+      "epoch": 226.16,
+      "grad_norm": 19.672819137573242,
+      "learning_rate": 2.4614614614614614e-06,
+      "loss": 0.3594,
+      "step": 75310
+    },
+    {
+      "epoch": 226.19,
+      "grad_norm": 15.298912048339844,
+      "learning_rate": 2.460460460460461e-06,
+      "loss": 0.3614,
+      "step": 75320
+    },
+    {
+      "epoch": 226.22,
+      "grad_norm": 18.152814865112305,
+      "learning_rate": 2.45945945945946e-06,
+      "loss": 0.3763,
+      "step": 75330
+    },
+    {
+      "epoch": 226.25,
+      "grad_norm": 13.384861946105957,
+      "learning_rate": 2.458458458458459e-06,
+      "loss": 0.3798,
+      "step": 75340
+    },
+    {
+      "epoch": 226.28,
+      "grad_norm": 21.438005447387695,
+      "learning_rate": 2.457457457457458e-06,
+      "loss": 0.3665,
+      "step": 75350
+    },
+    {
+      "epoch": 226.31,
+      "grad_norm": 16.457550048828125,
+      "learning_rate": 2.4564564564564564e-06,
+      "loss": 0.3426,
+      "step": 75360
+    },
+    {
+      "epoch": 226.34,
+      "grad_norm": 19.376127243041992,
+      "learning_rate": 2.455455455455456e-06,
+      "loss": 0.3679,
+      "step": 75370
+    },
+    {
+      "epoch": 226.37,
+      "grad_norm": 16.525136947631836,
+      "learning_rate": 2.4544544544544545e-06,
+      "loss": 0.29,
+      "step": 75380
+    },
+    {
+      "epoch": 226.4,
+      "grad_norm": 15.357576370239258,
+      "learning_rate": 2.453453453453454e-06,
+      "loss": 0.3621,
+      "step": 75390
+    },
+    {
+      "epoch": 226.43,
+      "grad_norm": 19.153520584106445,
+      "learning_rate": 2.4524524524524525e-06,
+      "loss": 0.3433,
+      "step": 75400
+    },
+    {
+      "epoch": 226.46,
+      "grad_norm": 18.370399475097656,
+      "learning_rate": 2.451451451451452e-06,
+      "loss": 0.3362,
+      "step": 75410
+    },
+    {
+      "epoch": 226.49,
+      "grad_norm": 20.147703170776367,
+      "learning_rate": 2.4504504504504506e-06,
+      "loss": 0.353,
+      "step": 75420
+    },
+    {
+      "epoch": 226.52,
+      "grad_norm": 22.971818923950195,
+      "learning_rate": 2.4494494494494496e-06,
+      "loss": 0.3201,
+      "step": 75430
+    },
+    {
+      "epoch": 226.55,
+      "grad_norm": 20.642709732055664,
+      "learning_rate": 2.4484484484484486e-06,
+      "loss": 0.3767,
+      "step": 75440
+    },
+    {
+      "epoch": 226.58,
+      "grad_norm": 27.393888473510742,
+      "learning_rate": 2.4474474474474476e-06,
+      "loss": 0.3108,
+      "step": 75450
+    },
+    {
+      "epoch": 226.61,
+      "grad_norm": 16.59218406677246,
+      "learning_rate": 2.4464464464464466e-06,
+      "loss": 0.3235,
+      "step": 75460
+    },
+    {
+      "epoch": 226.64,
+      "grad_norm": 13.26806926727295,
+      "learning_rate": 2.4454454454454456e-06,
+      "loss": 0.3036,
+      "step": 75470
+    },
+    {
+      "epoch": 226.67,
+      "grad_norm": 26.12692642211914,
+      "learning_rate": 2.4444444444444447e-06,
+      "loss": 0.3837,
+      "step": 75480
+    },
+    {
+      "epoch": 226.7,
+      "grad_norm": 23.029090881347656,
+      "learning_rate": 2.4434434434434437e-06,
+      "loss": 0.3256,
+      "step": 75490
+    },
+    {
+      "epoch": 226.73,
+      "grad_norm": 19.10053825378418,
+      "learning_rate": 2.4424424424424427e-06,
+      "loss": 0.3203,
+      "step": 75500
+    },
+    {
+      "epoch": 226.76,
+      "grad_norm": 22.952434539794922,
+      "learning_rate": 2.4414414414414417e-06,
+      "loss": 0.3163,
+      "step": 75510
+    },
+    {
+      "epoch": 226.79,
+      "grad_norm": 25.85211944580078,
+      "learning_rate": 2.4404404404404407e-06,
+      "loss": 0.3954,
+      "step": 75520
+    },
+    {
+      "epoch": 226.82,
+      "grad_norm": 13.743403434753418,
+      "learning_rate": 2.4394394394394398e-06,
+      "loss": 0.2916,
+      "step": 75530
+    },
+    {
+      "epoch": 226.85,
+      "grad_norm": 18.787067413330078,
+      "learning_rate": 2.4384384384384383e-06,
+      "loss": 0.3427,
+      "step": 75540
+    },
+    {
+      "epoch": 226.88,
+      "grad_norm": 11.99558162689209,
+      "learning_rate": 2.437437437437438e-06,
+      "loss": 0.322,
+      "step": 75550
+    },
+    {
+      "epoch": 226.91,
+      "grad_norm": 28.68096160888672,
+      "learning_rate": 2.4364364364364364e-06,
+      "loss": 0.3513,
+      "step": 75560
+    },
+    {
+      "epoch": 226.94,
+      "grad_norm": 16.49713134765625,
+      "learning_rate": 2.435435435435436e-06,
+      "loss": 0.3519,
+      "step": 75570
+    },
+    {
+      "epoch": 226.97,
+      "grad_norm": 25.362646102905273,
+      "learning_rate": 2.4344344344344344e-06,
+      "loss": 0.3424,
+      "step": 75580
+    },
+    {
+      "epoch": 227.0,
+      "grad_norm": 12.371591567993164,
+      "learning_rate": 2.4334334334334334e-06,
+      "loss": 0.327,
+      "step": 75590
+    },
+    {
+      "epoch": 227.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.5456568002700806,
+      "eval_runtime": 12.8626,
+      "eval_samples_per_second": 777.448,
+      "eval_steps_per_second": 3.11,
+      "step": 75591
+    },
+    {
+      "epoch": 227.03,
+      "grad_norm": 11.808660507202148,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.296,
+      "step": 75600
+    },
+    {
+      "epoch": 227.06,
+      "grad_norm": 18.831653594970703,
+      "learning_rate": 2.4314314314314315e-06,
+      "loss": 0.3012,
+      "step": 75610
+    },
+    {
+      "epoch": 227.09,
+      "grad_norm": 17.350296020507812,
+      "learning_rate": 2.430430430430431e-06,
+      "loss": 0.3718,
+      "step": 75620
+    },
+    {
+      "epoch": 227.12,
+      "grad_norm": 13.569050788879395,
+      "learning_rate": 2.4294294294294295e-06,
+      "loss": 0.3265,
+      "step": 75630
+    },
+    {
+      "epoch": 227.15,
+      "grad_norm": 13.02893352508545,
+      "learning_rate": 2.4284284284284285e-06,
+      "loss": 0.3633,
+      "step": 75640
+    },
+    {
+      "epoch": 227.18,
+      "grad_norm": 28.92671012878418,
+      "learning_rate": 2.4274274274274276e-06,
+      "loss": 0.3266,
+      "step": 75650
+    },
+    {
+      "epoch": 227.21,
+      "grad_norm": 13.801508903503418,
+      "learning_rate": 2.4264264264264266e-06,
+      "loss": 0.3842,
+      "step": 75660
+    },
+    {
+      "epoch": 227.24,
+      "grad_norm": 19.291080474853516,
+      "learning_rate": 2.4254254254254256e-06,
+      "loss": 0.3123,
+      "step": 75670
+    },
+    {
+      "epoch": 227.27,
+      "grad_norm": 18.59063148498535,
+      "learning_rate": 2.4244244244244246e-06,
+      "loss": 0.3592,
+      "step": 75680
+    },
+    {
+      "epoch": 227.3,
+      "grad_norm": 16.644121170043945,
+      "learning_rate": 2.4234234234234236e-06,
+      "loss": 0.321,
+      "step": 75690
+    },
+    {
+      "epoch": 227.33,
+      "grad_norm": 10.059311866760254,
+      "learning_rate": 2.4224224224224226e-06,
+      "loss": 0.3153,
+      "step": 75700
+    },
+    {
+      "epoch": 227.36,
+      "grad_norm": 14.858963012695312,
+      "learning_rate": 2.4214214214214217e-06,
+      "loss": 0.3368,
+      "step": 75710
+    },
+    {
+      "epoch": 227.39,
+      "grad_norm": 15.682539939880371,
+      "learning_rate": 2.4204204204204207e-06,
+      "loss": 0.36,
+      "step": 75720
+    },
+    {
+      "epoch": 227.42,
+      "grad_norm": 15.677034378051758,
+      "learning_rate": 2.4194194194194197e-06,
+      "loss": 0.3764,
+      "step": 75730
+    },
+    {
+      "epoch": 227.45,
+      "grad_norm": 17.585350036621094,
+      "learning_rate": 2.4184184184184187e-06,
+      "loss": 0.3298,
+      "step": 75740
+    },
+    {
+      "epoch": 227.48,
+      "grad_norm": 18.3528995513916,
+      "learning_rate": 2.4174174174174177e-06,
+      "loss": 0.375,
+      "step": 75750
+    },
+    {
+      "epoch": 227.51,
+      "grad_norm": 17.249528884887695,
+      "learning_rate": 2.4164164164164168e-06,
+      "loss": 0.3424,
+      "step": 75760
+    },
+    {
+      "epoch": 227.54,
+      "grad_norm": 17.440940856933594,
+      "learning_rate": 2.4154154154154153e-06,
+      "loss": 0.3572,
+      "step": 75770
+    },
+    {
+      "epoch": 227.57,
+      "grad_norm": 21.0602970123291,
+      "learning_rate": 2.414414414414415e-06,
+      "loss": 0.3609,
+      "step": 75780
+    },
+    {
+      "epoch": 227.6,
+      "grad_norm": 24.187225341796875,
+      "learning_rate": 2.4134134134134134e-06,
+      "loss": 0.3674,
+      "step": 75790
+    },
+    {
+      "epoch": 227.63,
+      "grad_norm": 10.687067985534668,
+      "learning_rate": 2.412412412412413e-06,
+      "loss": 0.3748,
+      "step": 75800
+    },
+    {
+      "epoch": 227.66,
+      "grad_norm": 15.824056625366211,
+      "learning_rate": 2.4114114114114114e-06,
+      "loss": 0.3979,
+      "step": 75810
+    },
+    {
+      "epoch": 227.69,
+      "grad_norm": 13.088824272155762,
+      "learning_rate": 2.4104104104104104e-06,
+      "loss": 0.294,
+      "step": 75820
+    },
+    {
+      "epoch": 227.72,
+      "grad_norm": 31.778608322143555,
+      "learning_rate": 2.4094094094094095e-06,
+      "loss": 0.3392,
+      "step": 75830
+    },
+    {
+      "epoch": 227.75,
+      "grad_norm": 14.002882957458496,
+      "learning_rate": 2.4084084084084085e-06,
+      "loss": 0.4034,
+      "step": 75840
+    },
+    {
+      "epoch": 227.78,
+      "grad_norm": 20.17993927001953,
+      "learning_rate": 2.4074074074074075e-06,
+      "loss": 0.3304,
+      "step": 75850
+    },
+    {
+      "epoch": 227.81,
+      "grad_norm": 16.854230880737305,
+      "learning_rate": 2.4064064064064065e-06,
+      "loss": 0.3534,
+      "step": 75860
+    },
+    {
+      "epoch": 227.84,
+      "grad_norm": 21.975303649902344,
+      "learning_rate": 2.4054054054054055e-06,
+      "loss": 0.3714,
+      "step": 75870
+    },
+    {
+      "epoch": 227.87,
+      "grad_norm": 20.86840057373047,
+      "learning_rate": 2.4044044044044045e-06,
+      "loss": 0.2803,
+      "step": 75880
+    },
+    {
+      "epoch": 227.9,
+      "grad_norm": 18.498252868652344,
+      "learning_rate": 2.4034034034034036e-06,
+      "loss": 0.3459,
+      "step": 75890
+    },
+    {
+      "epoch": 227.93,
+      "grad_norm": 10.389135360717773,
+      "learning_rate": 2.4024024024024026e-06,
+      "loss": 0.3187,
+      "step": 75900
+    },
+    {
+      "epoch": 227.96,
+      "grad_norm": 20.623836517333984,
+      "learning_rate": 2.4014014014014016e-06,
+      "loss": 0.3242,
+      "step": 75910
+    },
+    {
+      "epoch": 227.99,
+      "grad_norm": 14.08474063873291,
+      "learning_rate": 2.4004004004004006e-06,
+      "loss": 0.313,
+      "step": 75920
+    },
+    {
+      "epoch": 228.0,
+      "eval_accuracy": 0.8651,
+      "eval_loss": 0.5456980466842651,
+      "eval_runtime": 12.6631,
+      "eval_samples_per_second": 789.697,
+      "eval_steps_per_second": 3.159,
+      "step": 75924
+    },
+    {
+      "epoch": 228.02,
+      "grad_norm": 11.97934627532959,
+      "learning_rate": 2.3993993993993996e-06,
+      "loss": 0.35,
+      "step": 75930
+    },
+    {
+      "epoch": 228.05,
+      "grad_norm": 9.919551849365234,
+      "learning_rate": 2.3983983983983987e-06,
+      "loss": 0.4123,
+      "step": 75940
+    },
+    {
+      "epoch": 228.08,
+      "grad_norm": 25.207399368286133,
+      "learning_rate": 2.3973973973973977e-06,
+      "loss": 0.3038,
+      "step": 75950
+    },
+    {
+      "epoch": 228.11,
+      "grad_norm": 12.733241081237793,
+      "learning_rate": 2.3963963963963967e-06,
+      "loss": 0.3433,
+      "step": 75960
+    },
+    {
+      "epoch": 228.14,
+      "grad_norm": 15.249926567077637,
+      "learning_rate": 2.3953953953953957e-06,
+      "loss": 0.3214,
+      "step": 75970
+    },
+    {
+      "epoch": 228.17,
+      "grad_norm": 14.847517967224121,
+      "learning_rate": 2.3943943943943947e-06,
+      "loss": 0.3184,
+      "step": 75980
+    },
+    {
+      "epoch": 228.2,
+      "grad_norm": 16.630956649780273,
+      "learning_rate": 2.3933933933933938e-06,
+      "loss": 0.3572,
+      "step": 75990
+    },
+    {
+      "epoch": 228.23,
+      "grad_norm": 25.099042892456055,
+      "learning_rate": 2.3923923923923923e-06,
+      "loss": 0.3181,
+      "step": 76000
+    },
+    {
+      "epoch": 228.26,
+      "grad_norm": 15.082898139953613,
+      "learning_rate": 2.3913913913913918e-06,
+      "loss": 0.3417,
+      "step": 76010
+    },
+    {
+      "epoch": 228.29,
+      "grad_norm": 19.009504318237305,
+      "learning_rate": 2.3903903903903904e-06,
+      "loss": 0.339,
+      "step": 76020
+    },
+    {
+      "epoch": 228.32,
+      "grad_norm": 17.57427406311035,
+      "learning_rate": 2.38938938938939e-06,
+      "loss": 0.2798,
+      "step": 76030
+    },
+    {
+      "epoch": 228.35,
+      "grad_norm": 14.008241653442383,
+      "learning_rate": 2.3883883883883884e-06,
+      "loss": 0.328,
+      "step": 76040
+    },
+    {
+      "epoch": 228.38,
+      "grad_norm": 12.148763656616211,
+      "learning_rate": 2.3873873873873874e-06,
+      "loss": 0.3248,
+      "step": 76050
+    },
+    {
+      "epoch": 228.41,
+      "grad_norm": 13.519258499145508,
+      "learning_rate": 2.3863863863863865e-06,
+      "loss": 0.3146,
+      "step": 76060
+    },
+    {
+      "epoch": 228.44,
+      "grad_norm": 14.527026176452637,
+      "learning_rate": 2.3853853853853855e-06,
+      "loss": 0.3206,
+      "step": 76070
+    },
+    {
+      "epoch": 228.47,
+      "grad_norm": 25.646997451782227,
+      "learning_rate": 2.3843843843843845e-06,
+      "loss": 0.3125,
+      "step": 76080
+    },
+    {
+      "epoch": 228.5,
+      "grad_norm": 13.566756248474121,
+      "learning_rate": 2.3833833833833835e-06,
+      "loss": 0.4017,
+      "step": 76090
+    },
+    {
+      "epoch": 228.53,
+      "grad_norm": 48.38432312011719,
+      "learning_rate": 2.3823823823823825e-06,
+      "loss": 0.2986,
+      "step": 76100
+    },
+    {
+      "epoch": 228.56,
+      "grad_norm": 16.91596031188965,
+      "learning_rate": 2.3813813813813815e-06,
+      "loss": 0.322,
+      "step": 76110
+    },
+    {
+      "epoch": 228.59,
+      "grad_norm": 23.650806427001953,
+      "learning_rate": 2.3803803803803806e-06,
+      "loss": 0.3686,
+      "step": 76120
+    },
+    {
+      "epoch": 228.62,
+      "grad_norm": 12.536465644836426,
+      "learning_rate": 2.3793793793793796e-06,
+      "loss": 0.3486,
+      "step": 76130
+    },
+    {
+      "epoch": 228.65,
+      "grad_norm": 18.33787727355957,
+      "learning_rate": 2.3783783783783786e-06,
+      "loss": 0.3298,
+      "step": 76140
+    },
+    {
+      "epoch": 228.68,
+      "grad_norm": 19.01961898803711,
+      "learning_rate": 2.3773773773773776e-06,
+      "loss": 0.2935,
+      "step": 76150
+    },
+    {
+      "epoch": 228.71,
+      "grad_norm": 15.04861831665039,
+      "learning_rate": 2.3763763763763766e-06,
+      "loss": 0.2933,
+      "step": 76160
+    },
+    {
+      "epoch": 228.74,
+      "grad_norm": 14.101591110229492,
+      "learning_rate": 2.3753753753753757e-06,
+      "loss": 0.3136,
+      "step": 76170
+    },
+    {
+      "epoch": 228.77,
+      "grad_norm": 18.427358627319336,
+      "learning_rate": 2.3743743743743747e-06,
+      "loss": 0.3382,
+      "step": 76180
+    },
+    {
+      "epoch": 228.8,
+      "grad_norm": 14.729022026062012,
+      "learning_rate": 2.3733733733733737e-06,
+      "loss": 0.3094,
+      "step": 76190
+    },
+    {
+      "epoch": 228.83,
+      "grad_norm": 18.073780059814453,
+      "learning_rate": 2.3723723723723727e-06,
+      "loss": 0.3784,
+      "step": 76200
+    },
+    {
+      "epoch": 228.86,
+      "grad_norm": 20.68170928955078,
+      "learning_rate": 2.3713713713713717e-06,
+      "loss": 0.3517,
+      "step": 76210
+    },
+    {
+      "epoch": 228.89,
+      "grad_norm": 12.030599594116211,
+      "learning_rate": 2.3703703703703707e-06,
+      "loss": 0.3404,
+      "step": 76220
+    },
+    {
+      "epoch": 228.92,
+      "grad_norm": 16.608076095581055,
+      "learning_rate": 2.3693693693693693e-06,
+      "loss": 0.333,
+      "step": 76230
+    },
+    {
+      "epoch": 228.95,
+      "grad_norm": 16.127111434936523,
+      "learning_rate": 2.3683683683683688e-06,
+      "loss": 0.2988,
+      "step": 76240
+    },
+    {
+      "epoch": 228.98,
+      "grad_norm": 10.230194091796875,
+      "learning_rate": 2.3673673673673674e-06,
+      "loss": 0.3344,
+      "step": 76250
+    },
+    {
+      "epoch": 229.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5420966744422913,
+      "eval_runtime": 12.5276,
+      "eval_samples_per_second": 798.238,
+      "eval_steps_per_second": 3.193,
+      "step": 76257
+    },
+    {
+      "epoch": 229.01,
+      "grad_norm": 8.848485946655273,
+      "learning_rate": 2.366366366366367e-06,
+      "loss": 0.2812,
+      "step": 76260
+    },
+    {
+      "epoch": 229.04,
+      "grad_norm": 16.174488067626953,
+      "learning_rate": 2.3653653653653654e-06,
+      "loss": 0.3359,
+      "step": 76270
+    },
+    {
+      "epoch": 229.07,
+      "grad_norm": 13.698481559753418,
+      "learning_rate": 2.3643643643643644e-06,
+      "loss": 0.3524,
+      "step": 76280
+    },
+    {
+      "epoch": 229.1,
+      "grad_norm": 14.026788711547852,
+      "learning_rate": 2.3633633633633635e-06,
+      "loss": 0.3026,
+      "step": 76290
+    },
+    {
+      "epoch": 229.13,
+      "grad_norm": 12.95271110534668,
+      "learning_rate": 2.3623623623623625e-06,
+      "loss": 0.3676,
+      "step": 76300
+    },
+    {
+      "epoch": 229.16,
+      "grad_norm": 17.46615219116211,
+      "learning_rate": 2.3613613613613615e-06,
+      "loss": 0.2554,
+      "step": 76310
+    },
+    {
+      "epoch": 229.19,
+      "grad_norm": 17.75742530822754,
+      "learning_rate": 2.3603603603603605e-06,
+      "loss": 0.2986,
+      "step": 76320
+    },
+    {
+      "epoch": 229.22,
+      "grad_norm": 21.655540466308594,
+      "learning_rate": 2.3593593593593595e-06,
+      "loss": 0.3333,
+      "step": 76330
+    },
+    {
+      "epoch": 229.25,
+      "grad_norm": 22.020030975341797,
+      "learning_rate": 2.3583583583583585e-06,
+      "loss": 0.3112,
+      "step": 76340
+    },
+    {
+      "epoch": 229.28,
+      "grad_norm": 12.673250198364258,
+      "learning_rate": 2.3573573573573576e-06,
+      "loss": 0.3419,
+      "step": 76350
+    },
+    {
+      "epoch": 229.31,
+      "grad_norm": 17.323854446411133,
+      "learning_rate": 2.3563563563563566e-06,
+      "loss": 0.2976,
+      "step": 76360
+    },
+    {
+      "epoch": 229.34,
+      "grad_norm": 10.606208801269531,
+      "learning_rate": 2.3553553553553556e-06,
+      "loss": 0.3771,
+      "step": 76370
+    },
+    {
+      "epoch": 229.37,
+      "grad_norm": 23.976348876953125,
+      "learning_rate": 2.3543543543543546e-06,
+      "loss": 0.3562,
+      "step": 76380
+    },
+    {
+      "epoch": 229.4,
+      "grad_norm": 11.972336769104004,
+      "learning_rate": 2.3533533533533532e-06,
+      "loss": 0.2929,
+      "step": 76390
+    },
+    {
+      "epoch": 229.43,
+      "grad_norm": 19.057157516479492,
+      "learning_rate": 2.3523523523523527e-06,
+      "loss": 0.3487,
+      "step": 76400
+    },
+    {
+      "epoch": 229.46,
+      "grad_norm": 15.839333534240723,
+      "learning_rate": 2.3513513513513517e-06,
+      "loss": 0.3487,
+      "step": 76410
+    },
+    {
+      "epoch": 229.49,
+      "grad_norm": 24.80466079711914,
+      "learning_rate": 2.3503503503503507e-06,
+      "loss": 0.3139,
+      "step": 76420
+    },
+    {
+      "epoch": 229.52,
+      "grad_norm": 17.968544006347656,
+      "learning_rate": 2.3493493493493497e-06,
+      "loss": 0.3331,
+      "step": 76430
+    },
+    {
+      "epoch": 229.55,
+      "grad_norm": 16.602121353149414,
+      "learning_rate": 2.3483483483483483e-06,
+      "loss": 0.3361,
+      "step": 76440
+    },
+    {
+      "epoch": 229.58,
+      "grad_norm": 13.342581748962402,
+      "learning_rate": 2.3473473473473477e-06,
+      "loss": 0.3422,
+      "step": 76450
+    },
+    {
+      "epoch": 229.61,
+      "grad_norm": 20.721635818481445,
+      "learning_rate": 2.3463463463463463e-06,
+      "loss": 0.3509,
+      "step": 76460
+    },
+    {
+      "epoch": 229.64,
+      "grad_norm": 18.352264404296875,
+      "learning_rate": 2.3453453453453458e-06,
+      "loss": 0.3747,
+      "step": 76470
+    },
+    {
+      "epoch": 229.67,
+      "grad_norm": 10.96961498260498,
+      "learning_rate": 2.3443443443443444e-06,
+      "loss": 0.3494,
+      "step": 76480
+    },
+    {
+      "epoch": 229.7,
+      "grad_norm": 9.463398933410645,
+      "learning_rate": 2.343343343343344e-06,
+      "loss": 0.3081,
+      "step": 76490
+    },
+    {
+      "epoch": 229.73,
+      "grad_norm": 19.117300033569336,
+      "learning_rate": 2.3423423423423424e-06,
+      "loss": 0.3754,
+      "step": 76500
+    },
+    {
+      "epoch": 229.76,
+      "grad_norm": 12.680164337158203,
+      "learning_rate": 2.3413413413413414e-06,
+      "loss": 0.3345,
+      "step": 76510
+    },
+    {
+      "epoch": 229.79,
+      "grad_norm": 16.416845321655273,
+      "learning_rate": 2.3403403403403405e-06,
+      "loss": 0.3393,
+      "step": 76520
+    },
+    {
+      "epoch": 229.82,
+      "grad_norm": 14.378036499023438,
+      "learning_rate": 2.3393393393393395e-06,
+      "loss": 0.3049,
+      "step": 76530
+    },
+    {
+      "epoch": 229.85,
+      "grad_norm": 28.65460777282715,
+      "learning_rate": 2.3383383383383385e-06,
+      "loss": 0.3283,
+      "step": 76540
+    },
+    {
+      "epoch": 229.88,
+      "grad_norm": 11.441428184509277,
+      "learning_rate": 2.3373373373373375e-06,
+      "loss": 0.3087,
+      "step": 76550
+    },
+    {
+      "epoch": 229.91,
+      "grad_norm": 18.416227340698242,
+      "learning_rate": 2.3363363363363365e-06,
+      "loss": 0.2975,
+      "step": 76560
+    },
+    {
+      "epoch": 229.94,
+      "grad_norm": 14.096863746643066,
+      "learning_rate": 2.3353353353353355e-06,
+      "loss": 0.3537,
+      "step": 76570
+    },
+    {
+      "epoch": 229.97,
+      "grad_norm": 12.319059371948242,
+      "learning_rate": 2.3343343343343346e-06,
+      "loss": 0.423,
+      "step": 76580
+    },
+    {
+      "epoch": 230.0,
+      "grad_norm": 3.9853532314300537,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 0.2893,
+      "step": 76590
+    },
+    {
+      "epoch": 230.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5471625328063965,
+      "eval_runtime": 12.9316,
+      "eval_samples_per_second": 773.297,
+      "eval_steps_per_second": 3.093,
+      "step": 76590
+    },
+    {
+      "epoch": 230.03,
+      "grad_norm": 12.877554893493652,
+      "learning_rate": 2.3323323323323326e-06,
+      "loss": 0.344,
+      "step": 76600
+    },
+    {
+      "epoch": 230.06,
+      "grad_norm": 15.461255073547363,
+      "learning_rate": 2.3313313313313316e-06,
+      "loss": 0.3077,
+      "step": 76610
+    },
+    {
+      "epoch": 230.09,
+      "grad_norm": 16.680795669555664,
+      "learning_rate": 2.3303303303303302e-06,
+      "loss": 0.3757,
+      "step": 76620
+    },
+    {
+      "epoch": 230.12,
+      "grad_norm": 15.411928176879883,
+      "learning_rate": 2.3293293293293297e-06,
+      "loss": 0.3943,
+      "step": 76630
+    },
+    {
+      "epoch": 230.15,
+      "grad_norm": 16.059547424316406,
+      "learning_rate": 2.3283283283283282e-06,
+      "loss": 0.3336,
+      "step": 76640
+    },
+    {
+      "epoch": 230.18,
+      "grad_norm": 14.159675598144531,
+      "learning_rate": 2.3273273273273277e-06,
+      "loss": 0.3134,
+      "step": 76650
+    },
+    {
+      "epoch": 230.21,
+      "grad_norm": 15.191497802734375,
+      "learning_rate": 2.3263263263263263e-06,
+      "loss": 0.3271,
+      "step": 76660
+    },
+    {
+      "epoch": 230.24,
+      "grad_norm": 12.664679527282715,
+      "learning_rate": 2.3253253253253253e-06,
+      "loss": 0.313,
+      "step": 76670
+    },
+    {
+      "epoch": 230.27,
+      "grad_norm": 17.07796859741211,
+      "learning_rate": 2.3243243243243247e-06,
+      "loss": 0.3061,
+      "step": 76680
+    },
+    {
+      "epoch": 230.3,
+      "grad_norm": 12.30398178100586,
+      "learning_rate": 2.3233233233233233e-06,
+      "loss": 0.3065,
+      "step": 76690
+    },
+    {
+      "epoch": 230.33,
+      "grad_norm": 19.245389938354492,
+      "learning_rate": 2.3223223223223228e-06,
+      "loss": 0.384,
+      "step": 76700
+    },
+    {
+      "epoch": 230.36,
+      "grad_norm": 14.177104949951172,
+      "learning_rate": 2.3213213213213214e-06,
+      "loss": 0.3225,
+      "step": 76710
+    },
+    {
+      "epoch": 230.39,
+      "grad_norm": 13.848647117614746,
+      "learning_rate": 2.320320320320321e-06,
+      "loss": 0.3154,
+      "step": 76720
+    },
+    {
+      "epoch": 230.42,
+      "grad_norm": 16.30384635925293,
+      "learning_rate": 2.3193193193193194e-06,
+      "loss": 0.3487,
+      "step": 76730
+    },
+    {
+      "epoch": 230.45,
+      "grad_norm": 12.804328918457031,
+      "learning_rate": 2.3183183183183184e-06,
+      "loss": 0.3448,
+      "step": 76740
+    },
+    {
+      "epoch": 230.48,
+      "grad_norm": 18.986791610717773,
+      "learning_rate": 2.3173173173173175e-06,
+      "loss": 0.3295,
+      "step": 76750
+    },
+    {
+      "epoch": 230.51,
+      "grad_norm": 17.67308807373047,
+      "learning_rate": 2.3163163163163165e-06,
+      "loss": 0.3297,
+      "step": 76760
+    },
+    {
+      "epoch": 230.54,
+      "grad_norm": 14.166356086730957,
+      "learning_rate": 2.3153153153153155e-06,
+      "loss": 0.3064,
+      "step": 76770
+    },
+    {
+      "epoch": 230.57,
+      "grad_norm": 12.981823921203613,
+      "learning_rate": 2.3143143143143145e-06,
+      "loss": 0.33,
+      "step": 76780
+    },
+    {
+      "epoch": 230.6,
+      "grad_norm": 27.787403106689453,
+      "learning_rate": 2.3133133133133135e-06,
+      "loss": 0.3453,
+      "step": 76790
+    },
+    {
+      "epoch": 230.63,
+      "grad_norm": 16.26099967956543,
+      "learning_rate": 2.3123123123123125e-06,
+      "loss": 0.3433,
+      "step": 76800
+    },
+    {
+      "epoch": 230.66,
+      "grad_norm": 11.284934043884277,
+      "learning_rate": 2.3113113113113116e-06,
+      "loss": 0.3576,
+      "step": 76810
+    },
+    {
+      "epoch": 230.69,
+      "grad_norm": 11.245282173156738,
+      "learning_rate": 2.3103103103103106e-06,
+      "loss": 0.364,
+      "step": 76820
+    },
+    {
+      "epoch": 230.72,
+      "grad_norm": 11.966828346252441,
+      "learning_rate": 2.3093093093093096e-06,
+      "loss": 0.3844,
+      "step": 76830
+    },
+    {
+      "epoch": 230.75,
+      "grad_norm": 14.267149925231934,
+      "learning_rate": 2.3083083083083086e-06,
+      "loss": 0.3761,
+      "step": 76840
+    },
+    {
+      "epoch": 230.78,
+      "grad_norm": 25.82526397705078,
+      "learning_rate": 2.307307307307307e-06,
+      "loss": 0.3383,
+      "step": 76850
+    },
+    {
+      "epoch": 230.81,
+      "grad_norm": 13.665443420410156,
+      "learning_rate": 2.3063063063063067e-06,
+      "loss": 0.3699,
+      "step": 76860
+    },
+    {
+      "epoch": 230.84,
+      "grad_norm": 18.922792434692383,
+      "learning_rate": 2.3053053053053052e-06,
+      "loss": 0.3685,
+      "step": 76870
+    },
+    {
+      "epoch": 230.87,
+      "grad_norm": 16.672592163085938,
+      "learning_rate": 2.3043043043043047e-06,
+      "loss": 0.3486,
+      "step": 76880
+    },
+    {
+      "epoch": 230.9,
+      "grad_norm": 19.249427795410156,
+      "learning_rate": 2.3033033033033033e-06,
+      "loss": 0.3547,
+      "step": 76890
+    },
+    {
+      "epoch": 230.93,
+      "grad_norm": 20.18529510498047,
+      "learning_rate": 2.3023023023023023e-06,
+      "loss": 0.362,
+      "step": 76900
+    },
+    {
+      "epoch": 230.96,
+      "grad_norm": 16.873140335083008,
+      "learning_rate": 2.3013013013013013e-06,
+      "loss": 0.3234,
+      "step": 76910
+    },
+    {
+      "epoch": 230.99,
+      "grad_norm": 10.984187126159668,
+      "learning_rate": 2.3003003003003003e-06,
+      "loss": 0.3225,
+      "step": 76920
+    },
+    {
+      "epoch": 231.0,
+      "eval_accuracy": 0.8651,
+      "eval_loss": 0.5436414480209351,
+      "eval_runtime": 12.9426,
+      "eval_samples_per_second": 772.643,
+      "eval_steps_per_second": 3.091,
+      "step": 76923
+    },
+    {
+      "epoch": 231.02,
+      "grad_norm": 21.69346809387207,
+      "learning_rate": 2.2992992992992998e-06,
+      "loss": 0.2864,
+      "step": 76930
+    },
+    {
+      "epoch": 231.05,
+      "grad_norm": 11.366324424743652,
+      "learning_rate": 2.2982982982982984e-06,
+      "loss": 0.3198,
+      "step": 76940
+    },
+    {
+      "epoch": 231.08,
+      "grad_norm": 18.07667350769043,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 0.3527,
+      "step": 76950
+    },
+    {
+      "epoch": 231.11,
+      "grad_norm": 14.143817901611328,
+      "learning_rate": 2.2962962962962964e-06,
+      "loss": 0.3049,
+      "step": 76960
+    },
+    {
+      "epoch": 231.14,
+      "grad_norm": 20.538841247558594,
+      "learning_rate": 2.2952952952952954e-06,
+      "loss": 0.2825,
+      "step": 76970
+    },
+    {
+      "epoch": 231.17,
+      "grad_norm": 13.512066841125488,
+      "learning_rate": 2.2942942942942944e-06,
+      "loss": 0.3436,
+      "step": 76980
+    },
+    {
+      "epoch": 231.2,
+      "grad_norm": 21.66156768798828,
+      "learning_rate": 2.2932932932932935e-06,
+      "loss": 0.333,
+      "step": 76990
+    },
+    {
+      "epoch": 231.23,
+      "grad_norm": 13.490944862365723,
+      "learning_rate": 2.2922922922922925e-06,
+      "loss": 0.3155,
+      "step": 77000
+    },
+    {
+      "epoch": 231.26,
+      "grad_norm": 16.140026092529297,
+      "learning_rate": 2.2912912912912915e-06,
+      "loss": 0.3307,
+      "step": 77010
+    },
+    {
+      "epoch": 231.29,
+      "grad_norm": 17.864805221557617,
+      "learning_rate": 2.2902902902902905e-06,
+      "loss": 0.3466,
+      "step": 77020
+    },
+    {
+      "epoch": 231.32,
+      "grad_norm": 15.11877727508545,
+      "learning_rate": 2.2892892892892895e-06,
+      "loss": 0.3493,
+      "step": 77030
+    },
+    {
+      "epoch": 231.35,
+      "grad_norm": 7.7971367835998535,
+      "learning_rate": 2.2882882882882886e-06,
+      "loss": 0.3638,
+      "step": 77040
+    },
+    {
+      "epoch": 231.38,
+      "grad_norm": 18.067405700683594,
+      "learning_rate": 2.2872872872872876e-06,
+      "loss": 0.3622,
+      "step": 77050
+    },
+    {
+      "epoch": 231.41,
+      "grad_norm": 16.68392562866211,
+      "learning_rate": 2.2862862862862866e-06,
+      "loss": 0.2876,
+      "step": 77060
+    },
+    {
+      "epoch": 231.44,
+      "grad_norm": 13.168264389038086,
+      "learning_rate": 2.2852852852852856e-06,
+      "loss": 0.3239,
+      "step": 77070
+    },
+    {
+      "epoch": 231.47,
+      "grad_norm": 12.142965316772461,
+      "learning_rate": 2.284284284284284e-06,
+      "loss": 0.2889,
+      "step": 77080
+    },
+    {
+      "epoch": 231.5,
+      "grad_norm": 16.00080680847168,
+      "learning_rate": 2.2832832832832837e-06,
+      "loss": 0.3041,
+      "step": 77090
+    },
+    {
+      "epoch": 231.53,
+      "grad_norm": 18.455291748046875,
+      "learning_rate": 2.2822822822822822e-06,
+      "loss": 0.3558,
+      "step": 77100
+    },
+    {
+      "epoch": 231.56,
+      "grad_norm": 15.978042602539062,
+      "learning_rate": 2.2812812812812817e-06,
+      "loss": 0.3736,
+      "step": 77110
+    },
+    {
+      "epoch": 231.59,
+      "grad_norm": 20.255935668945312,
+      "learning_rate": 2.2802802802802803e-06,
+      "loss": 0.3478,
+      "step": 77120
+    },
+    {
+      "epoch": 231.62,
+      "grad_norm": 11.982696533203125,
+      "learning_rate": 2.2792792792792793e-06,
+      "loss": 0.3466,
+      "step": 77130
+    },
+    {
+      "epoch": 231.65,
+      "grad_norm": 20.942710876464844,
+      "learning_rate": 2.2782782782782783e-06,
+      "loss": 0.3015,
+      "step": 77140
+    },
+    {
+      "epoch": 231.68,
+      "grad_norm": 13.490928649902344,
+      "learning_rate": 2.2772772772772773e-06,
+      "loss": 0.2718,
+      "step": 77150
+    },
+    {
+      "epoch": 231.71,
+      "grad_norm": 12.858659744262695,
+      "learning_rate": 2.2762762762762764e-06,
+      "loss": 0.3261,
+      "step": 77160
+    },
+    {
+      "epoch": 231.74,
+      "grad_norm": 14.599413871765137,
+      "learning_rate": 2.2752752752752754e-06,
+      "loss": 0.3781,
+      "step": 77170
+    },
+    {
+      "epoch": 231.77,
+      "grad_norm": 18.732458114624023,
+      "learning_rate": 2.2742742742742744e-06,
+      "loss": 0.3733,
+      "step": 77180
+    },
+    {
+      "epoch": 231.8,
+      "grad_norm": 15.306767463684082,
+      "learning_rate": 2.2732732732732734e-06,
+      "loss": 0.3761,
+      "step": 77190
+    },
+    {
+      "epoch": 231.83,
+      "grad_norm": 12.832221984863281,
+      "learning_rate": 2.2722722722722724e-06,
+      "loss": 0.2916,
+      "step": 77200
+    },
+    {
+      "epoch": 231.86,
+      "grad_norm": 18.923683166503906,
+      "learning_rate": 2.2712712712712714e-06,
+      "loss": 0.3261,
+      "step": 77210
+    },
+    {
+      "epoch": 231.89,
+      "grad_norm": 18.551319122314453,
+      "learning_rate": 2.2702702702702705e-06,
+      "loss": 0.3529,
+      "step": 77220
+    },
+    {
+      "epoch": 231.92,
+      "grad_norm": 19.12047576904297,
+      "learning_rate": 2.2692692692692695e-06,
+      "loss": 0.3707,
+      "step": 77230
+    },
+    {
+      "epoch": 231.95,
+      "grad_norm": 8.545114517211914,
+      "learning_rate": 2.2682682682682685e-06,
+      "loss": 0.306,
+      "step": 77240
+    },
+    {
+      "epoch": 231.98,
+      "grad_norm": 20.494287490844727,
+      "learning_rate": 2.2672672672672675e-06,
+      "loss": 0.3662,
+      "step": 77250
+    },
+    {
+      "epoch": 232.0,
+      "eval_accuracy": 0.8654,
+      "eval_loss": 0.5428200960159302,
+      "eval_runtime": 13.0157,
+      "eval_samples_per_second": 768.302,
+      "eval_steps_per_second": 3.073,
+      "step": 77256
+    },
+    {
+      "epoch": 232.01,
+      "grad_norm": 17.20171546936035,
+      "learning_rate": 2.2662662662662665e-06,
+      "loss": 0.3443,
+      "step": 77260
+    },
+    {
+      "epoch": 232.04,
+      "grad_norm": 37.830291748046875,
+      "learning_rate": 2.2652652652652656e-06,
+      "loss": 0.3361,
+      "step": 77270
+    },
+    {
+      "epoch": 232.07,
+      "grad_norm": 21.27008056640625,
+      "learning_rate": 2.2642642642642646e-06,
+      "loss": 0.363,
+      "step": 77280
+    },
+    {
+      "epoch": 232.1,
+      "grad_norm": 15.784087181091309,
+      "learning_rate": 2.2632632632632636e-06,
+      "loss": 0.3558,
+      "step": 77290
+    },
+    {
+      "epoch": 232.13,
+      "grad_norm": 20.176889419555664,
+      "learning_rate": 2.2622622622622626e-06,
+      "loss": 0.3906,
+      "step": 77300
+    },
+    {
+      "epoch": 232.16,
+      "grad_norm": 16.504985809326172,
+      "learning_rate": 2.261261261261261e-06,
+      "loss": 0.3488,
+      "step": 77310
+    },
+    {
+      "epoch": 232.19,
+      "grad_norm": 12.659575462341309,
+      "learning_rate": 2.2602602602602606e-06,
+      "loss": 0.3076,
+      "step": 77320
+    },
+    {
+      "epoch": 232.22,
+      "grad_norm": 11.724394798278809,
+      "learning_rate": 2.2592592592592592e-06,
+      "loss": 0.3305,
+      "step": 77330
+    },
+    {
+      "epoch": 232.25,
+      "grad_norm": 13.674764633178711,
+      "learning_rate": 2.2582582582582587e-06,
+      "loss": 0.3146,
+      "step": 77340
+    },
+    {
+      "epoch": 232.28,
+      "grad_norm": 18.7500057220459,
+      "learning_rate": 2.2572572572572573e-06,
+      "loss": 0.2886,
+      "step": 77350
+    },
+    {
+      "epoch": 232.31,
+      "grad_norm": 18.563974380493164,
+      "learning_rate": 2.2562562562562563e-06,
+      "loss": 0.4069,
+      "step": 77360
+    },
+    {
+      "epoch": 232.34,
+      "grad_norm": 14.002694129943848,
+      "learning_rate": 2.2552552552552553e-06,
+      "loss": 0.322,
+      "step": 77370
+    },
+    {
+      "epoch": 232.37,
+      "grad_norm": 15.807002067565918,
+      "learning_rate": 2.2542542542542543e-06,
+      "loss": 0.3514,
+      "step": 77380
+    },
+    {
+      "epoch": 232.4,
+      "grad_norm": 13.782280921936035,
+      "learning_rate": 2.2532532532532534e-06,
+      "loss": 0.3241,
+      "step": 77390
+    },
+    {
+      "epoch": 232.43,
+      "grad_norm": 15.79702377319336,
+      "learning_rate": 2.2522522522522524e-06,
+      "loss": 0.3265,
+      "step": 77400
+    },
+    {
+      "epoch": 232.46,
+      "grad_norm": 15.372262954711914,
+      "learning_rate": 2.2512512512512514e-06,
+      "loss": 0.3827,
+      "step": 77410
+    },
+    {
+      "epoch": 232.49,
+      "grad_norm": 16.113964080810547,
+      "learning_rate": 2.2502502502502504e-06,
+      "loss": 0.331,
+      "step": 77420
+    },
+    {
+      "epoch": 232.52,
+      "grad_norm": 16.159902572631836,
+      "learning_rate": 2.2492492492492494e-06,
+      "loss": 0.4183,
+      "step": 77430
+    },
+    {
+      "epoch": 232.55,
+      "grad_norm": 20.522809982299805,
+      "learning_rate": 2.2482482482482484e-06,
+      "loss": 0.2988,
+      "step": 77440
+    },
+    {
+      "epoch": 232.58,
+      "grad_norm": 21.798280715942383,
+      "learning_rate": 2.2472472472472475e-06,
+      "loss": 0.3315,
+      "step": 77450
+    },
+    {
+      "epoch": 232.61,
+      "grad_norm": 14.840535163879395,
+      "learning_rate": 2.2462462462462465e-06,
+      "loss": 0.341,
+      "step": 77460
+    },
+    {
+      "epoch": 232.64,
+      "grad_norm": 12.696000099182129,
+      "learning_rate": 2.2452452452452455e-06,
+      "loss": 0.3393,
+      "step": 77470
+    },
+    {
+      "epoch": 232.67,
+      "grad_norm": 18.317907333374023,
+      "learning_rate": 2.2442442442442445e-06,
+      "loss": 0.3084,
+      "step": 77480
+    },
+    {
+      "epoch": 232.7,
+      "grad_norm": 17.257444381713867,
+      "learning_rate": 2.2432432432432435e-06,
+      "loss": 0.3129,
+      "step": 77490
+    },
+    {
+      "epoch": 232.73,
+      "grad_norm": 13.81556224822998,
+      "learning_rate": 2.2422422422422426e-06,
+      "loss": 0.2956,
+      "step": 77500
+    },
+    {
+      "epoch": 232.76,
+      "grad_norm": 17.331199645996094,
+      "learning_rate": 2.2412412412412416e-06,
+      "loss": 0.3428,
+      "step": 77510
+    },
+    {
+      "epoch": 232.79,
+      "grad_norm": 13.60155200958252,
+      "learning_rate": 2.2402402402402406e-06,
+      "loss": 0.4009,
+      "step": 77520
+    },
+    {
+      "epoch": 232.82,
+      "grad_norm": 19.039718627929688,
+      "learning_rate": 2.2392392392392396e-06,
+      "loss": 0.3631,
+      "step": 77530
+    },
+    {
+      "epoch": 232.85,
+      "grad_norm": 18.60478401184082,
+      "learning_rate": 2.238238238238238e-06,
+      "loss": 0.3552,
+      "step": 77540
+    },
+    {
+      "epoch": 232.88,
+      "grad_norm": 12.313214302062988,
+      "learning_rate": 2.2372372372372376e-06,
+      "loss": 0.3831,
+      "step": 77550
+    },
+    {
+      "epoch": 232.91,
+      "grad_norm": 13.185256004333496,
+      "learning_rate": 2.2362362362362362e-06,
+      "loss": 0.3248,
+      "step": 77560
+    },
+    {
+      "epoch": 232.94,
+      "grad_norm": 19.992355346679688,
+      "learning_rate": 2.2352352352352357e-06,
+      "loss": 0.3296,
+      "step": 77570
+    },
+    {
+      "epoch": 232.97,
+      "grad_norm": 18.987985610961914,
+      "learning_rate": 2.2342342342342343e-06,
+      "loss": 0.3281,
+      "step": 77580
+    },
+    {
+      "epoch": 233.0,
+      "eval_accuracy": 0.8654,
+      "eval_loss": 0.5453376173973083,
+      "eval_runtime": 12.8569,
+      "eval_samples_per_second": 777.793,
+      "eval_steps_per_second": 3.111,
+      "step": 77589
+    },
+    {
+      "epoch": 233.0,
+      "grad_norm": 29.624597549438477,
+      "learning_rate": 2.2332332332332333e-06,
+      "loss": 0.333,
+      "step": 77590
+    },
+    {
+      "epoch": 233.03,
+      "grad_norm": 15.667610168457031,
+      "learning_rate": 2.2322322322322323e-06,
+      "loss": 0.3476,
+      "step": 77600
+    },
+    {
+      "epoch": 233.06,
+      "grad_norm": 17.561185836791992,
+      "learning_rate": 2.2312312312312313e-06,
+      "loss": 0.3037,
+      "step": 77610
+    },
+    {
+      "epoch": 233.09,
+      "grad_norm": 17.28537940979004,
+      "learning_rate": 2.2302302302302304e-06,
+      "loss": 0.3221,
+      "step": 77620
+    },
+    {
+      "epoch": 233.12,
+      "grad_norm": 16.238847732543945,
+      "learning_rate": 2.2292292292292294e-06,
+      "loss": 0.312,
+      "step": 77630
+    },
+    {
+      "epoch": 233.15,
+      "grad_norm": 12.182740211486816,
+      "learning_rate": 2.2282282282282284e-06,
+      "loss": 0.3322,
+      "step": 77640
+    },
+    {
+      "epoch": 233.18,
+      "grad_norm": 12.298293113708496,
+      "learning_rate": 2.2272272272272274e-06,
+      "loss": 0.3335,
+      "step": 77650
+    },
+    {
+      "epoch": 233.21,
+      "grad_norm": 15.505232810974121,
+      "learning_rate": 2.2262262262262264e-06,
+      "loss": 0.3452,
+      "step": 77660
+    },
+    {
+      "epoch": 233.24,
+      "grad_norm": 13.163336753845215,
+      "learning_rate": 2.2252252252252254e-06,
+      "loss": 0.3425,
+      "step": 77670
+    },
+    {
+      "epoch": 233.27,
+      "grad_norm": 13.627297401428223,
+      "learning_rate": 2.2242242242242245e-06,
+      "loss": 0.2977,
+      "step": 77680
+    },
+    {
+      "epoch": 233.3,
+      "grad_norm": 16.533632278442383,
+      "learning_rate": 2.2232232232232235e-06,
+      "loss": 0.3284,
+      "step": 77690
+    },
+    {
+      "epoch": 233.33,
+      "grad_norm": 21.957548141479492,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.3112,
+      "step": 77700
+    },
+    {
+      "epoch": 233.36,
+      "grad_norm": 23.161571502685547,
+      "learning_rate": 2.2212212212212215e-06,
+      "loss": 0.3436,
+      "step": 77710
+    },
+    {
+      "epoch": 233.39,
+      "grad_norm": 26.680763244628906,
+      "learning_rate": 2.22022022022022e-06,
+      "loss": 0.4303,
+      "step": 77720
+    },
+    {
+      "epoch": 233.42,
+      "grad_norm": 16.87899398803711,
+      "learning_rate": 2.2192192192192196e-06,
+      "loss": 0.3308,
+      "step": 77730
+    },
+    {
+      "epoch": 233.45,
+      "grad_norm": 13.929515838623047,
+      "learning_rate": 2.2182182182182186e-06,
+      "loss": 0.3717,
+      "step": 77740
+    },
+    {
+      "epoch": 233.48,
+      "grad_norm": 17.483049392700195,
+      "learning_rate": 2.2172172172172176e-06,
+      "loss": 0.3341,
+      "step": 77750
+    },
+    {
+      "epoch": 233.51,
+      "grad_norm": 16.270965576171875,
+      "learning_rate": 2.2162162162162166e-06,
+      "loss": 0.381,
+      "step": 77760
+    },
+    {
+      "epoch": 233.54,
+      "grad_norm": 18.02756690979004,
+      "learning_rate": 2.215215215215215e-06,
+      "loss": 0.3315,
+      "step": 77770
+    },
+    {
+      "epoch": 233.57,
+      "grad_norm": 32.32350158691406,
+      "learning_rate": 2.2142142142142146e-06,
+      "loss": 0.3361,
+      "step": 77780
+    },
+    {
+      "epoch": 233.6,
+      "grad_norm": 15.346711158752441,
+      "learning_rate": 2.2132132132132132e-06,
+      "loss": 0.3067,
+      "step": 77790
+    },
+    {
+      "epoch": 233.63,
+      "grad_norm": 25.037166595458984,
+      "learning_rate": 2.2122122122122127e-06,
+      "loss": 0.2958,
+      "step": 77800
+    },
+    {
+      "epoch": 233.66,
+      "grad_norm": 21.041784286499023,
+      "learning_rate": 2.2112112112112113e-06,
+      "loss": 0.3244,
+      "step": 77810
+    },
+    {
+      "epoch": 233.69,
+      "grad_norm": 19.80297088623047,
+      "learning_rate": 2.2102102102102103e-06,
+      "loss": 0.3115,
+      "step": 77820
+    },
+    {
+      "epoch": 233.72,
+      "grad_norm": 16.87148666381836,
+      "learning_rate": 2.2092092092092093e-06,
+      "loss": 0.3605,
+      "step": 77830
+    },
+    {
+      "epoch": 233.75,
+      "grad_norm": 14.597328186035156,
+      "learning_rate": 2.2082082082082083e-06,
+      "loss": 0.3565,
+      "step": 77840
+    },
+    {
+      "epoch": 233.78,
+      "grad_norm": 14.174348831176758,
+      "learning_rate": 2.2072072072072073e-06,
+      "loss": 0.3444,
+      "step": 77850
+    },
+    {
+      "epoch": 233.81,
+      "grad_norm": 21.404399871826172,
+      "learning_rate": 2.2062062062062064e-06,
+      "loss": 0.2822,
+      "step": 77860
+    },
+    {
+      "epoch": 233.84,
+      "grad_norm": 15.5905122756958,
+      "learning_rate": 2.2052052052052054e-06,
+      "loss": 0.3525,
+      "step": 77870
+    },
+    {
+      "epoch": 233.87,
+      "grad_norm": 18.9205265045166,
+      "learning_rate": 2.2042042042042044e-06,
+      "loss": 0.3107,
+      "step": 77880
+    },
+    {
+      "epoch": 233.9,
+      "grad_norm": 16.737104415893555,
+      "learning_rate": 2.2032032032032034e-06,
+      "loss": 0.3532,
+      "step": 77890
+    },
+    {
+      "epoch": 233.93,
+      "grad_norm": 14.232735633850098,
+      "learning_rate": 2.2022022022022024e-06,
+      "loss": 0.3223,
+      "step": 77900
+    },
+    {
+      "epoch": 233.96,
+      "grad_norm": 11.80434513092041,
+      "learning_rate": 2.2012012012012015e-06,
+      "loss": 0.2808,
+      "step": 77910
+    },
+    {
+      "epoch": 233.99,
+      "grad_norm": 17.103519439697266,
+      "learning_rate": 2.2002002002002005e-06,
+      "loss": 0.3354,
+      "step": 77920
+    },
+    {
+      "epoch": 234.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5467805862426758,
+      "eval_runtime": 12.8403,
+      "eval_samples_per_second": 778.796,
+      "eval_steps_per_second": 3.115,
+      "step": 77922
+    },
+    {
+      "epoch": 234.02,
+      "grad_norm": 12.720157623291016,
+      "learning_rate": 2.199199199199199e-06,
+      "loss": 0.3026,
+      "step": 77930
+    },
+    {
+      "epoch": 234.05,
+      "grad_norm": 9.480108261108398,
+      "learning_rate": 2.1981981981981985e-06,
+      "loss": 0.3213,
+      "step": 77940
+    },
+    {
+      "epoch": 234.08,
+      "grad_norm": 22.482328414916992,
+      "learning_rate": 2.197197197197197e-06,
+      "loss": 0.3839,
+      "step": 77950
+    },
+    {
+      "epoch": 234.11,
+      "grad_norm": 13.445293426513672,
+      "learning_rate": 2.1961961961961966e-06,
+      "loss": 0.3031,
+      "step": 77960
+    },
+    {
+      "epoch": 234.14,
+      "grad_norm": 15.00788688659668,
+      "learning_rate": 2.195195195195195e-06,
+      "loss": 0.3707,
+      "step": 77970
+    },
+    {
+      "epoch": 234.17,
+      "grad_norm": 9.617106437683105,
+      "learning_rate": 2.1941941941941946e-06,
+      "loss": 0.3185,
+      "step": 77980
+    },
+    {
+      "epoch": 234.2,
+      "grad_norm": 12.985164642333984,
+      "learning_rate": 2.193193193193193e-06,
+      "loss": 0.289,
+      "step": 77990
+    },
+    {
+      "epoch": 234.23,
+      "grad_norm": 10.963327407836914,
+      "learning_rate": 2.192192192192192e-06,
+      "loss": 0.3205,
+      "step": 78000
+    },
+    {
+      "epoch": 234.26,
+      "grad_norm": 33.009281158447266,
+      "learning_rate": 2.1911911911911916e-06,
+      "loss": 0.3846,
+      "step": 78010
+    },
+    {
+      "epoch": 234.29,
+      "grad_norm": 22.02430534362793,
+      "learning_rate": 2.1901901901901902e-06,
+      "loss": 0.318,
+      "step": 78020
+    },
+    {
+      "epoch": 234.32,
+      "grad_norm": 13.009856224060059,
+      "learning_rate": 2.1891891891891897e-06,
+      "loss": 0.2703,
+      "step": 78030
+    },
+    {
+      "epoch": 234.35,
+      "grad_norm": 16.42791175842285,
+      "learning_rate": 2.1881881881881883e-06,
+      "loss": 0.3099,
+      "step": 78040
+    },
+    {
+      "epoch": 234.38,
+      "grad_norm": 17.300458908081055,
+      "learning_rate": 2.1871871871871873e-06,
+      "loss": 0.3651,
+      "step": 78050
+    },
+    {
+      "epoch": 234.41,
+      "grad_norm": 15.60451602935791,
+      "learning_rate": 2.1861861861861863e-06,
+      "loss": 0.3223,
+      "step": 78060
+    },
+    {
+      "epoch": 234.44,
+      "grad_norm": 22.71717071533203,
+      "learning_rate": 2.1851851851851853e-06,
+      "loss": 0.3467,
+      "step": 78070
+    },
+    {
+      "epoch": 234.47,
+      "grad_norm": 22.459461212158203,
+      "learning_rate": 2.1841841841841843e-06,
+      "loss": 0.3362,
+      "step": 78080
+    },
+    {
+      "epoch": 234.5,
+      "grad_norm": 17.568363189697266,
+      "learning_rate": 2.1831831831831834e-06,
+      "loss": 0.3618,
+      "step": 78090
+    },
+    {
+      "epoch": 234.53,
+      "grad_norm": 9.771783828735352,
+      "learning_rate": 2.1821821821821824e-06,
+      "loss": 0.3004,
+      "step": 78100
+    },
+    {
+      "epoch": 234.56,
+      "grad_norm": 22.78858757019043,
+      "learning_rate": 2.1811811811811814e-06,
+      "loss": 0.3311,
+      "step": 78110
+    },
+    {
+      "epoch": 234.59,
+      "grad_norm": 15.481013298034668,
+      "learning_rate": 2.1801801801801804e-06,
+      "loss": 0.3687,
+      "step": 78120
+    },
+    {
+      "epoch": 234.62,
+      "grad_norm": 19.270689010620117,
+      "learning_rate": 2.1791791791791794e-06,
+      "loss": 0.3572,
+      "step": 78130
+    },
+    {
+      "epoch": 234.65,
+      "grad_norm": 26.37474250793457,
+      "learning_rate": 2.1781781781781785e-06,
+      "loss": 0.3502,
+      "step": 78140
+    },
+    {
+      "epoch": 234.68,
+      "grad_norm": 17.595245361328125,
+      "learning_rate": 2.1771771771771775e-06,
+      "loss": 0.317,
+      "step": 78150
+    },
+    {
+      "epoch": 234.71,
+      "grad_norm": 15.045089721679688,
+      "learning_rate": 2.176176176176176e-06,
+      "loss": 0.3168,
+      "step": 78160
+    },
+    {
+      "epoch": 234.74,
+      "grad_norm": 12.883399963378906,
+      "learning_rate": 2.1751751751751755e-06,
+      "loss": 0.3068,
+      "step": 78170
+    },
+    {
+      "epoch": 234.77,
+      "grad_norm": 15.119956016540527,
+      "learning_rate": 2.174174174174174e-06,
+      "loss": 0.3465,
+      "step": 78180
+    },
+    {
+      "epoch": 234.8,
+      "grad_norm": 16.62995147705078,
+      "learning_rate": 2.1731731731731735e-06,
+      "loss": 0.3913,
+      "step": 78190
+    },
+    {
+      "epoch": 234.83,
+      "grad_norm": 17.205677032470703,
+      "learning_rate": 2.172172172172172e-06,
+      "loss": 0.3245,
+      "step": 78200
+    },
+    {
+      "epoch": 234.86,
+      "grad_norm": 20.600679397583008,
+      "learning_rate": 2.1711711711711716e-06,
+      "loss": 0.3665,
+      "step": 78210
+    },
+    {
+      "epoch": 234.89,
+      "grad_norm": 17.06092643737793,
+      "learning_rate": 2.17017017017017e-06,
+      "loss": 0.3531,
+      "step": 78220
+    },
+    {
+      "epoch": 234.92,
+      "grad_norm": 25.04252815246582,
+      "learning_rate": 2.169169169169169e-06,
+      "loss": 0.366,
+      "step": 78230
+    },
+    {
+      "epoch": 234.95,
+      "grad_norm": 18.571125030517578,
+      "learning_rate": 2.1681681681681682e-06,
+      "loss": 0.3338,
+      "step": 78240
+    },
+    {
+      "epoch": 234.98,
+      "grad_norm": 13.873452186584473,
+      "learning_rate": 2.1671671671671672e-06,
+      "loss": 0.3238,
+      "step": 78250
+    },
+    {
+      "epoch": 235.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5501341223716736,
+      "eval_runtime": 12.5884,
+      "eval_samples_per_second": 794.382,
+      "eval_steps_per_second": 3.178,
+      "step": 78255
+    },
+    {
+      "epoch": 235.02,
+      "grad_norm": 15.202035903930664,
+      "learning_rate": 2.1661661661661663e-06,
+      "loss": 0.4033,
+      "step": 78260
+    },
+    {
+      "epoch": 235.05,
+      "grad_norm": 13.207056999206543,
+      "learning_rate": 2.1651651651651653e-06,
+      "loss": 0.3073,
+      "step": 78270
+    },
+    {
+      "epoch": 235.08,
+      "grad_norm": 16.15162467956543,
+      "learning_rate": 2.1641641641641643e-06,
+      "loss": 0.2893,
+      "step": 78280
+    },
+    {
+      "epoch": 235.11,
+      "grad_norm": 15.350812911987305,
+      "learning_rate": 2.1631631631631633e-06,
+      "loss": 0.3468,
+      "step": 78290
+    },
+    {
+      "epoch": 235.14,
+      "grad_norm": 16.368623733520508,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.3457,
+      "step": 78300
+    },
+    {
+      "epoch": 235.17,
+      "grad_norm": 19.685012817382812,
+      "learning_rate": 2.1611611611611613e-06,
+      "loss": 0.3024,
+      "step": 78310
+    },
+    {
+      "epoch": 235.2,
+      "grad_norm": 12.172863006591797,
+      "learning_rate": 2.1601601601601604e-06,
+      "loss": 0.3473,
+      "step": 78320
+    },
+    {
+      "epoch": 235.23,
+      "grad_norm": 13.041714668273926,
+      "learning_rate": 2.1591591591591594e-06,
+      "loss": 0.2988,
+      "step": 78330
+    },
+    {
+      "epoch": 235.26,
+      "grad_norm": 15.117774963378906,
+      "learning_rate": 2.1581581581581584e-06,
+      "loss": 0.3313,
+      "step": 78340
+    },
+    {
+      "epoch": 235.29,
+      "grad_norm": 18.15601348876953,
+      "learning_rate": 2.1571571571571574e-06,
+      "loss": 0.3482,
+      "step": 78350
+    },
+    {
+      "epoch": 235.32,
+      "grad_norm": 14.645926475524902,
+      "learning_rate": 2.1561561561561564e-06,
+      "loss": 0.3453,
+      "step": 78360
+    },
+    {
+      "epoch": 235.35,
+      "grad_norm": 11.506360054016113,
+      "learning_rate": 2.1551551551551555e-06,
+      "loss": 0.3156,
+      "step": 78370
+    },
+    {
+      "epoch": 235.38,
+      "grad_norm": 16.800765991210938,
+      "learning_rate": 2.1541541541541545e-06,
+      "loss": 0.2874,
+      "step": 78380
+    },
+    {
+      "epoch": 235.41,
+      "grad_norm": 13.600683212280273,
+      "learning_rate": 2.153153153153153e-06,
+      "loss": 0.3422,
+      "step": 78390
+    },
+    {
+      "epoch": 235.44,
+      "grad_norm": 10.64301586151123,
+      "learning_rate": 2.1521521521521525e-06,
+      "loss": 0.3163,
+      "step": 78400
+    },
+    {
+      "epoch": 235.47,
+      "grad_norm": 13.463558197021484,
+      "learning_rate": 2.151151151151151e-06,
+      "loss": 0.3208,
+      "step": 78410
+    },
+    {
+      "epoch": 235.5,
+      "grad_norm": 14.028066635131836,
+      "learning_rate": 2.1501501501501505e-06,
+      "loss": 0.3416,
+      "step": 78420
+    },
+    {
+      "epoch": 235.53,
+      "grad_norm": 15.132709503173828,
+      "learning_rate": 2.149149149149149e-06,
+      "loss": 0.353,
+      "step": 78430
+    },
+    {
+      "epoch": 235.56,
+      "grad_norm": 13.419441223144531,
+      "learning_rate": 2.148148148148148e-06,
+      "loss": 0.3334,
+      "step": 78440
+    },
+    {
+      "epoch": 235.59,
+      "grad_norm": 15.435487747192383,
+      "learning_rate": 2.147147147147147e-06,
+      "loss": 0.3576,
+      "step": 78450
+    },
+    {
+      "epoch": 235.62,
+      "grad_norm": 17.936420440673828,
+      "learning_rate": 2.146146146146146e-06,
+      "loss": 0.3123,
+      "step": 78460
+    },
+    {
+      "epoch": 235.65,
+      "grad_norm": 18.038942337036133,
+      "learning_rate": 2.1451451451451452e-06,
+      "loss": 0.355,
+      "step": 78470
+    },
+    {
+      "epoch": 235.68,
+      "grad_norm": 19.990428924560547,
+      "learning_rate": 2.1441441441441442e-06,
+      "loss": 0.3997,
+      "step": 78480
+    },
+    {
+      "epoch": 235.71,
+      "grad_norm": 16.307846069335938,
+      "learning_rate": 2.1431431431431433e-06,
+      "loss": 0.3253,
+      "step": 78490
+    },
+    {
+      "epoch": 235.74,
+      "grad_norm": 8.991814613342285,
+      "learning_rate": 2.1421421421421423e-06,
+      "loss": 0.3195,
+      "step": 78500
+    },
+    {
+      "epoch": 235.77,
+      "grad_norm": 31.374984741210938,
+      "learning_rate": 2.1411411411411413e-06,
+      "loss": 0.3242,
+      "step": 78510
+    },
+    {
+      "epoch": 235.8,
+      "grad_norm": 18.197175979614258,
+      "learning_rate": 2.1401401401401403e-06,
+      "loss": 0.3401,
+      "step": 78520
+    },
+    {
+      "epoch": 235.83,
+      "grad_norm": 24.70032501220703,
+      "learning_rate": 2.1391391391391393e-06,
+      "loss": 0.3266,
+      "step": 78530
+    },
+    {
+      "epoch": 235.86,
+      "grad_norm": 18.22015953063965,
+      "learning_rate": 2.1381381381381383e-06,
+      "loss": 0.3335,
+      "step": 78540
+    },
+    {
+      "epoch": 235.89,
+      "grad_norm": 15.12339973449707,
+      "learning_rate": 2.1371371371371374e-06,
+      "loss": 0.3816,
+      "step": 78550
+    },
+    {
+      "epoch": 235.92,
+      "grad_norm": 19.205215454101562,
+      "learning_rate": 2.1361361361361364e-06,
+      "loss": 0.3327,
+      "step": 78560
+    },
+    {
+      "epoch": 235.95,
+      "grad_norm": 17.1258487701416,
+      "learning_rate": 2.1351351351351354e-06,
+      "loss": 0.3292,
+      "step": 78570
+    },
+    {
+      "epoch": 235.98,
+      "grad_norm": 21.429086685180664,
+      "learning_rate": 2.1341341341341344e-06,
+      "loss": 0.292,
+      "step": 78580
+    },
+    {
+      "epoch": 236.0,
+      "eval_accuracy": 0.8658,
+      "eval_loss": 0.5419029593467712,
+      "eval_runtime": 12.88,
+      "eval_samples_per_second": 776.397,
+      "eval_steps_per_second": 3.106,
+      "step": 78588
+    },
+    {
+      "epoch": 236.01,
+      "grad_norm": 17.396865844726562,
+      "learning_rate": 2.1331331331331334e-06,
+      "loss": 0.3252,
+      "step": 78590
+    },
+    {
+      "epoch": 236.04,
+      "grad_norm": 14.123645782470703,
+      "learning_rate": 2.1321321321321325e-06,
+      "loss": 0.3156,
+      "step": 78600
+    },
+    {
+      "epoch": 236.07,
+      "grad_norm": 17.886676788330078,
+      "learning_rate": 2.1311311311311315e-06,
+      "loss": 0.34,
+      "step": 78610
+    },
+    {
+      "epoch": 236.1,
+      "grad_norm": 18.64170265197754,
+      "learning_rate": 2.13013013013013e-06,
+      "loss": 0.3561,
+      "step": 78620
+    },
+    {
+      "epoch": 236.13,
+      "grad_norm": 16.759262084960938,
+      "learning_rate": 2.1291291291291295e-06,
+      "loss": 0.3504,
+      "step": 78630
+    },
+    {
+      "epoch": 236.16,
+      "grad_norm": 11.85936450958252,
+      "learning_rate": 2.128128128128128e-06,
+      "loss": 0.305,
+      "step": 78640
+    },
+    {
+      "epoch": 236.19,
+      "grad_norm": 19.407621383666992,
+      "learning_rate": 2.1271271271271275e-06,
+      "loss": 0.3401,
+      "step": 78650
+    },
+    {
+      "epoch": 236.22,
+      "grad_norm": 16.955148696899414,
+      "learning_rate": 2.126126126126126e-06,
+      "loss": 0.3382,
+      "step": 78660
+    },
+    {
+      "epoch": 236.25,
+      "grad_norm": 15.628632545471191,
+      "learning_rate": 2.125125125125125e-06,
+      "loss": 0.3269,
+      "step": 78670
+    },
+    {
+      "epoch": 236.28,
+      "grad_norm": 22.925045013427734,
+      "learning_rate": 2.124124124124124e-06,
+      "loss": 0.3644,
+      "step": 78680
+    },
+    {
+      "epoch": 236.31,
+      "grad_norm": 14.206666946411133,
+      "learning_rate": 2.123123123123123e-06,
+      "loss": 0.3201,
+      "step": 78690
+    },
+    {
+      "epoch": 236.34,
+      "grad_norm": 19.20172882080078,
+      "learning_rate": 2.1221221221221222e-06,
+      "loss": 0.3315,
+      "step": 78700
+    },
+    {
+      "epoch": 236.37,
+      "grad_norm": 17.937820434570312,
+      "learning_rate": 2.1211211211211212e-06,
+      "loss": 0.3321,
+      "step": 78710
+    },
+    {
+      "epoch": 236.4,
+      "grad_norm": 15.272162437438965,
+      "learning_rate": 2.1201201201201203e-06,
+      "loss": 0.3597,
+      "step": 78720
+    },
+    {
+      "epoch": 236.43,
+      "grad_norm": 18.070201873779297,
+      "learning_rate": 2.1191191191191193e-06,
+      "loss": 0.3732,
+      "step": 78730
+    },
+    {
+      "epoch": 236.46,
+      "grad_norm": 19.057830810546875,
+      "learning_rate": 2.1181181181181183e-06,
+      "loss": 0.345,
+      "step": 78740
+    },
+    {
+      "epoch": 236.49,
+      "grad_norm": 12.185697555541992,
+      "learning_rate": 2.1171171171171173e-06,
+      "loss": 0.3291,
+      "step": 78750
+    },
+    {
+      "epoch": 236.52,
+      "grad_norm": 12.540802955627441,
+      "learning_rate": 2.1161161161161163e-06,
+      "loss": 0.3502,
+      "step": 78760
+    },
+    {
+      "epoch": 236.55,
+      "grad_norm": 12.832952499389648,
+      "learning_rate": 2.1151151151151153e-06,
+      "loss": 0.323,
+      "step": 78770
+    },
+    {
+      "epoch": 236.58,
+      "grad_norm": 11.47408390045166,
+      "learning_rate": 2.1141141141141144e-06,
+      "loss": 0.3335,
+      "step": 78780
+    },
+    {
+      "epoch": 236.61,
+      "grad_norm": 10.238903045654297,
+      "learning_rate": 2.1131131131131134e-06,
+      "loss": 0.3095,
+      "step": 78790
+    },
+    {
+      "epoch": 236.64,
+      "grad_norm": 13.602645874023438,
+      "learning_rate": 2.1121121121121124e-06,
+      "loss": 0.3218,
+      "step": 78800
+    },
+    {
+      "epoch": 236.67,
+      "grad_norm": 15.575439453125,
+      "learning_rate": 2.1111111111111114e-06,
+      "loss": 0.3452,
+      "step": 78810
+    },
+    {
+      "epoch": 236.7,
+      "grad_norm": 14.182901382446289,
+      "learning_rate": 2.1101101101101104e-06,
+      "loss": 0.354,
+      "step": 78820
+    },
+    {
+      "epoch": 236.73,
+      "grad_norm": 24.669631958007812,
+      "learning_rate": 2.1091091091091095e-06,
+      "loss": 0.3893,
+      "step": 78830
+    },
+    {
+      "epoch": 236.76,
+      "grad_norm": 13.30527400970459,
+      "learning_rate": 2.1081081081081085e-06,
+      "loss": 0.3508,
+      "step": 78840
+    },
+    {
+      "epoch": 236.79,
+      "grad_norm": 21.09571647644043,
+      "learning_rate": 2.107107107107107e-06,
+      "loss": 0.4104,
+      "step": 78850
+    },
+    {
+      "epoch": 236.82,
+      "grad_norm": 18.163930892944336,
+      "learning_rate": 2.1061061061061065e-06,
+      "loss": 0.3204,
+      "step": 78860
+    },
+    {
+      "epoch": 236.85,
+      "grad_norm": 11.39368724822998,
+      "learning_rate": 2.105105105105105e-06,
+      "loss": 0.3328,
+      "step": 78870
+    },
+    {
+      "epoch": 236.88,
+      "grad_norm": 15.885597229003906,
+      "learning_rate": 2.1041041041041045e-06,
+      "loss": 0.2896,
+      "step": 78880
+    },
+    {
+      "epoch": 236.91,
+      "grad_norm": 13.3842134475708,
+      "learning_rate": 2.103103103103103e-06,
+      "loss": 0.3507,
+      "step": 78890
+    },
+    {
+      "epoch": 236.94,
+      "grad_norm": 10.832818031311035,
+      "learning_rate": 2.102102102102102e-06,
+      "loss": 0.3481,
+      "step": 78900
+    },
+    {
+      "epoch": 236.97,
+      "grad_norm": 16.01152992248535,
+      "learning_rate": 2.101101101101101e-06,
+      "loss": 0.3498,
+      "step": 78910
+    },
+    {
+      "epoch": 237.0,
+      "grad_norm": 26.154449462890625,
+      "learning_rate": 2.1001001001001e-06,
+      "loss": 0.3863,
+      "step": 78920
+    },
+    {
+      "epoch": 237.0,
+      "eval_accuracy": 0.8637,
+      "eval_loss": 0.5445350408554077,
+      "eval_runtime": 12.7499,
+      "eval_samples_per_second": 784.321,
+      "eval_steps_per_second": 3.137,
+      "step": 78921
+    },
+    {
+      "epoch": 237.03,
+      "grad_norm": 23.513904571533203,
+      "learning_rate": 2.0990990990990992e-06,
+      "loss": 0.347,
+      "step": 78930
+    },
+    {
+      "epoch": 237.06,
+      "grad_norm": 13.018270492553711,
+      "learning_rate": 2.0980980980980982e-06,
+      "loss": 0.3516,
+      "step": 78940
+    },
+    {
+      "epoch": 237.09,
+      "grad_norm": 15.532323837280273,
+      "learning_rate": 2.0970970970970972e-06,
+      "loss": 0.3439,
+      "step": 78950
+    },
+    {
+      "epoch": 237.12,
+      "grad_norm": 12.503837585449219,
+      "learning_rate": 2.0960960960960963e-06,
+      "loss": 0.3662,
+      "step": 78960
+    },
+    {
+      "epoch": 237.15,
+      "grad_norm": 22.816619873046875,
+      "learning_rate": 2.0950950950950953e-06,
+      "loss": 0.3365,
+      "step": 78970
+    },
+    {
+      "epoch": 237.18,
+      "grad_norm": 14.084867477416992,
+      "learning_rate": 2.0940940940940943e-06,
+      "loss": 0.3373,
+      "step": 78980
+    },
+    {
+      "epoch": 237.21,
+      "grad_norm": 18.695018768310547,
+      "learning_rate": 2.0930930930930933e-06,
+      "loss": 0.3246,
+      "step": 78990
+    },
+    {
+      "epoch": 237.24,
+      "grad_norm": 15.444843292236328,
+      "learning_rate": 2.0920920920920923e-06,
+      "loss": 0.3477,
+      "step": 79000
+    },
+    {
+      "epoch": 237.27,
+      "grad_norm": 30.305559158325195,
+      "learning_rate": 2.0910910910910914e-06,
+      "loss": 0.3817,
+      "step": 79010
+    },
+    {
+      "epoch": 237.3,
+      "grad_norm": 9.856484413146973,
+      "learning_rate": 2.0900900900900904e-06,
+      "loss": 0.3258,
+      "step": 79020
+    },
+    {
+      "epoch": 237.33,
+      "grad_norm": 16.019184112548828,
+      "learning_rate": 2.089089089089089e-06,
+      "loss": 0.3553,
+      "step": 79030
+    },
+    {
+      "epoch": 237.36,
+      "grad_norm": 19.802003860473633,
+      "learning_rate": 2.0880880880880884e-06,
+      "loss": 0.3324,
+      "step": 79040
+    },
+    {
+      "epoch": 237.39,
+      "grad_norm": 12.311623573303223,
+      "learning_rate": 2.087087087087087e-06,
+      "loss": 0.3349,
+      "step": 79050
+    },
+    {
+      "epoch": 237.42,
+      "grad_norm": 16.898780822753906,
+      "learning_rate": 2.0860860860860865e-06,
+      "loss": 0.3291,
+      "step": 79060
+    },
+    {
+      "epoch": 237.45,
+      "grad_norm": 16.056955337524414,
+      "learning_rate": 2.0850850850850855e-06,
+      "loss": 0.318,
+      "step": 79070
+    },
+    {
+      "epoch": 237.48,
+      "grad_norm": 10.872939109802246,
+      "learning_rate": 2.084084084084084e-06,
+      "loss": 0.3103,
+      "step": 79080
+    },
+    {
+      "epoch": 237.51,
+      "grad_norm": 21.182470321655273,
+      "learning_rate": 2.0830830830830835e-06,
+      "loss": 0.3487,
+      "step": 79090
+    },
+    {
+      "epoch": 237.54,
+      "grad_norm": 13.969568252563477,
+      "learning_rate": 2.082082082082082e-06,
+      "loss": 0.3309,
+      "step": 79100
+    },
+    {
+      "epoch": 237.57,
+      "grad_norm": 15.308022499084473,
+      "learning_rate": 2.0810810810810815e-06,
+      "loss": 0.3442,
+      "step": 79110
+    },
+    {
+      "epoch": 237.6,
+      "grad_norm": 16.8736629486084,
+      "learning_rate": 2.08008008008008e-06,
+      "loss": 0.348,
+      "step": 79120
+    },
+    {
+      "epoch": 237.63,
+      "grad_norm": 16.510149002075195,
+      "learning_rate": 2.079079079079079e-06,
+      "loss": 0.3158,
+      "step": 79130
+    },
+    {
+      "epoch": 237.66,
+      "grad_norm": 18.71401596069336,
+      "learning_rate": 2.078078078078078e-06,
+      "loss": 0.3714,
+      "step": 79140
+    },
+    {
+      "epoch": 237.69,
+      "grad_norm": 18.421836853027344,
+      "learning_rate": 2.077077077077077e-06,
+      "loss": 0.368,
+      "step": 79150
+    },
+    {
+      "epoch": 237.72,
+      "grad_norm": 12.561145782470703,
+      "learning_rate": 2.076076076076076e-06,
+      "loss": 0.3706,
+      "step": 79160
+    },
+    {
+      "epoch": 237.75,
+      "grad_norm": 13.893301963806152,
+      "learning_rate": 2.0750750750750752e-06,
+      "loss": 0.3188,
+      "step": 79170
+    },
+    {
+      "epoch": 237.78,
+      "grad_norm": 14.794346809387207,
+      "learning_rate": 2.0740740740740742e-06,
+      "loss": 0.3276,
+      "step": 79180
+    },
+    {
+      "epoch": 237.81,
+      "grad_norm": 14.895833015441895,
+      "learning_rate": 2.0730730730730733e-06,
+      "loss": 0.3228,
+      "step": 79190
+    },
+    {
+      "epoch": 237.84,
+      "grad_norm": 14.927363395690918,
+      "learning_rate": 2.0720720720720723e-06,
+      "loss": 0.3574,
+      "step": 79200
+    },
+    {
+      "epoch": 237.87,
+      "grad_norm": 20.943004608154297,
+      "learning_rate": 2.0710710710710713e-06,
+      "loss": 0.3001,
+      "step": 79210
+    },
+    {
+      "epoch": 237.9,
+      "grad_norm": 17.045475006103516,
+      "learning_rate": 2.0700700700700703e-06,
+      "loss": 0.348,
+      "step": 79220
+    },
+    {
+      "epoch": 237.93,
+      "grad_norm": 23.346115112304688,
+      "learning_rate": 2.0690690690690693e-06,
+      "loss": 0.3811,
+      "step": 79230
+    },
+    {
+      "epoch": 237.96,
+      "grad_norm": 16.207027435302734,
+      "learning_rate": 2.0680680680680684e-06,
+      "loss": 0.3357,
+      "step": 79240
+    },
+    {
+      "epoch": 237.99,
+      "grad_norm": 24.359909057617188,
+      "learning_rate": 2.0670670670670674e-06,
+      "loss": 0.3368,
+      "step": 79250
+    },
+    {
+      "epoch": 238.0,
+      "eval_accuracy": 0.8643,
+      "eval_loss": 0.5450624227523804,
+      "eval_runtime": 12.6705,
+      "eval_samples_per_second": 789.234,
+      "eval_steps_per_second": 3.157,
+      "step": 79254
+    },
+    {
+      "epoch": 238.02,
+      "grad_norm": 11.002306938171387,
+      "learning_rate": 2.066066066066066e-06,
+      "loss": 0.3128,
+      "step": 79260
+    },
+    {
+      "epoch": 238.05,
+      "grad_norm": 11.958415985107422,
+      "learning_rate": 2.0650650650650654e-06,
+      "loss": 0.311,
+      "step": 79270
+    },
+    {
+      "epoch": 238.08,
+      "grad_norm": 16.814441680908203,
+      "learning_rate": 2.064064064064064e-06,
+      "loss": 0.339,
+      "step": 79280
+    },
+    {
+      "epoch": 238.11,
+      "grad_norm": 15.511554718017578,
+      "learning_rate": 2.0630630630630634e-06,
+      "loss": 0.3285,
+      "step": 79290
+    },
+    {
+      "epoch": 238.14,
+      "grad_norm": 22.84657859802246,
+      "learning_rate": 2.062062062062062e-06,
+      "loss": 0.3404,
+      "step": 79300
+    },
+    {
+      "epoch": 238.17,
+      "grad_norm": 17.35125732421875,
+      "learning_rate": 2.061061061061061e-06,
+      "loss": 0.3322,
+      "step": 79310
+    },
+    {
+      "epoch": 238.2,
+      "grad_norm": 16.070350646972656,
+      "learning_rate": 2.06006006006006e-06,
+      "loss": 0.3249,
+      "step": 79320
+    },
+    {
+      "epoch": 238.23,
+      "grad_norm": 13.452096939086914,
+      "learning_rate": 2.059059059059059e-06,
+      "loss": 0.3479,
+      "step": 79330
+    },
+    {
+      "epoch": 238.26,
+      "grad_norm": 10.907296180725098,
+      "learning_rate": 2.0580580580580585e-06,
+      "loss": 0.3389,
+      "step": 79340
+    },
+    {
+      "epoch": 238.29,
+      "grad_norm": 22.135873794555664,
+      "learning_rate": 2.057057057057057e-06,
+      "loss": 0.328,
+      "step": 79350
+    },
+    {
+      "epoch": 238.32,
+      "grad_norm": 10.63663101196289,
+      "learning_rate": 2.056056056056056e-06,
+      "loss": 0.3475,
+      "step": 79360
+    },
+    {
+      "epoch": 238.35,
+      "grad_norm": 13.685857772827148,
+      "learning_rate": 2.055055055055055e-06,
+      "loss": 0.318,
+      "step": 79370
+    },
+    {
+      "epoch": 238.38,
+      "grad_norm": 12.421021461486816,
+      "learning_rate": 2.054054054054054e-06,
+      "loss": 0.3272,
+      "step": 79380
+    },
+    {
+      "epoch": 238.41,
+      "grad_norm": 13.2546968460083,
+      "learning_rate": 2.053053053053053e-06,
+      "loss": 0.3157,
+      "step": 79390
+    },
+    {
+      "epoch": 238.44,
+      "grad_norm": 17.05144500732422,
+      "learning_rate": 2.0520520520520522e-06,
+      "loss": 0.3687,
+      "step": 79400
+    },
+    {
+      "epoch": 238.47,
+      "grad_norm": 16.936376571655273,
+      "learning_rate": 2.0510510510510512e-06,
+      "loss": 0.3734,
+      "step": 79410
+    },
+    {
+      "epoch": 238.5,
+      "grad_norm": 14.262466430664062,
+      "learning_rate": 2.0500500500500503e-06,
+      "loss": 0.3646,
+      "step": 79420
+    },
+    {
+      "epoch": 238.53,
+      "grad_norm": 18.57318878173828,
+      "learning_rate": 2.0490490490490493e-06,
+      "loss": 0.305,
+      "step": 79430
+    },
+    {
+      "epoch": 238.56,
+      "grad_norm": 13.315982818603516,
+      "learning_rate": 2.0480480480480483e-06,
+      "loss": 0.3538,
+      "step": 79440
+    },
+    {
+      "epoch": 238.59,
+      "grad_norm": 13.153244972229004,
+      "learning_rate": 2.0470470470470473e-06,
+      "loss": 0.3413,
+      "step": 79450
+    },
+    {
+      "epoch": 238.62,
+      "grad_norm": 13.260665893554688,
+      "learning_rate": 2.0460460460460463e-06,
+      "loss": 0.3395,
+      "step": 79460
+    },
+    {
+      "epoch": 238.65,
+      "grad_norm": 11.628409385681152,
+      "learning_rate": 2.045045045045045e-06,
+      "loss": 0.3036,
+      "step": 79470
+    },
+    {
+      "epoch": 238.68,
+      "grad_norm": 13.154693603515625,
+      "learning_rate": 2.0440440440440444e-06,
+      "loss": 0.3023,
+      "step": 79480
+    },
+    {
+      "epoch": 238.71,
+      "grad_norm": 18.033411026000977,
+      "learning_rate": 2.043043043043043e-06,
+      "loss": 0.3762,
+      "step": 79490
+    },
+    {
+      "epoch": 238.74,
+      "grad_norm": 28.11608123779297,
+      "learning_rate": 2.0420420420420424e-06,
+      "loss": 0.4157,
+      "step": 79500
+    },
+    {
+      "epoch": 238.77,
+      "grad_norm": 12.814567565917969,
+      "learning_rate": 2.041041041041041e-06,
+      "loss": 0.3458,
+      "step": 79510
+    },
+    {
+      "epoch": 238.8,
+      "grad_norm": 12.712309837341309,
+      "learning_rate": 2.0400400400400404e-06,
+      "loss": 0.3539,
+      "step": 79520
+    },
+    {
+      "epoch": 238.83,
+      "grad_norm": 18.168128967285156,
+      "learning_rate": 2.039039039039039e-06,
+      "loss": 0.3172,
+      "step": 79530
+    },
+    {
+      "epoch": 238.86,
+      "grad_norm": 9.311602592468262,
+      "learning_rate": 2.038038038038038e-06,
+      "loss": 0.3059,
+      "step": 79540
+    },
+    {
+      "epoch": 238.89,
+      "grad_norm": 9.41295051574707,
+      "learning_rate": 2.037037037037037e-06,
+      "loss": 0.3306,
+      "step": 79550
+    },
+    {
+      "epoch": 238.92,
+      "grad_norm": 17.544097900390625,
+      "learning_rate": 2.036036036036036e-06,
+      "loss": 0.328,
+      "step": 79560
+    },
+    {
+      "epoch": 238.95,
+      "grad_norm": 18.023509979248047,
+      "learning_rate": 2.035035035035035e-06,
+      "loss": 0.2985,
+      "step": 79570
+    },
+    {
+      "epoch": 238.98,
+      "grad_norm": 10.872513771057129,
+      "learning_rate": 2.034034034034034e-06,
+      "loss": 0.3011,
+      "step": 79580
+    },
+    {
+      "epoch": 239.0,
+      "eval_accuracy": 0.8651,
+      "eval_loss": 0.5458724498748779,
+      "eval_runtime": 12.8115,
+      "eval_samples_per_second": 780.548,
+      "eval_steps_per_second": 3.122,
+      "step": 79587
+    },
+    {
+      "epoch": 239.01,
+      "grad_norm": 11.73851203918457,
+      "learning_rate": 2.033033033033033e-06,
+      "loss": 0.3334,
+      "step": 79590
+    },
+    {
+      "epoch": 239.04,
+      "grad_norm": 16.847091674804688,
+      "learning_rate": 2.032032032032032e-06,
+      "loss": 0.3573,
+      "step": 79600
+    },
+    {
+      "epoch": 239.07,
+      "grad_norm": 17.49502944946289,
+      "learning_rate": 2.031031031031031e-06,
+      "loss": 0.3521,
+      "step": 79610
+    },
+    {
+      "epoch": 239.1,
+      "grad_norm": 21.909818649291992,
+      "learning_rate": 2.03003003003003e-06,
+      "loss": 0.2935,
+      "step": 79620
+    },
+    {
+      "epoch": 239.13,
+      "grad_norm": 16.76329231262207,
+      "learning_rate": 2.0290290290290292e-06,
+      "loss": 0.3917,
+      "step": 79630
+    },
+    {
+      "epoch": 239.16,
+      "grad_norm": 15.793485641479492,
+      "learning_rate": 2.0280280280280282e-06,
+      "loss": 0.3054,
+      "step": 79640
+    },
+    {
+      "epoch": 239.19,
+      "grad_norm": 10.483351707458496,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.3532,
+      "step": 79650
+    },
+    {
+      "epoch": 239.22,
+      "grad_norm": 13.077339172363281,
+      "learning_rate": 2.0260260260260263e-06,
+      "loss": 0.3335,
+      "step": 79660
+    },
+    {
+      "epoch": 239.25,
+      "grad_norm": 15.774230003356934,
+      "learning_rate": 2.0250250250250253e-06,
+      "loss": 0.3229,
+      "step": 79670
+    },
+    {
+      "epoch": 239.28,
+      "grad_norm": 19.05290412902832,
+      "learning_rate": 2.0240240240240243e-06,
+      "loss": 0.3701,
+      "step": 79680
+    },
+    {
+      "epoch": 239.31,
+      "grad_norm": 16.707748413085938,
+      "learning_rate": 2.0230230230230233e-06,
+      "loss": 0.3082,
+      "step": 79690
+    },
+    {
+      "epoch": 239.34,
+      "grad_norm": 12.222843170166016,
+      "learning_rate": 2.022022022022022e-06,
+      "loss": 0.3682,
+      "step": 79700
+    },
+    {
+      "epoch": 239.37,
+      "grad_norm": 8.958308219909668,
+      "learning_rate": 2.0210210210210214e-06,
+      "loss": 0.3624,
+      "step": 79710
+    },
+    {
+      "epoch": 239.4,
+      "grad_norm": 13.62440299987793,
+      "learning_rate": 2.02002002002002e-06,
+      "loss": 0.3323,
+      "step": 79720
+    },
+    {
+      "epoch": 239.43,
+      "grad_norm": 8.738441467285156,
+      "learning_rate": 2.0190190190190194e-06,
+      "loss": 0.3154,
+      "step": 79730
+    },
+    {
+      "epoch": 239.46,
+      "grad_norm": 19.11353302001953,
+      "learning_rate": 2.018018018018018e-06,
+      "loss": 0.3298,
+      "step": 79740
+    },
+    {
+      "epoch": 239.49,
+      "grad_norm": 13.462930679321289,
+      "learning_rate": 2.0170170170170174e-06,
+      "loss": 0.3593,
+      "step": 79750
+    },
+    {
+      "epoch": 239.52,
+      "grad_norm": 13.082551956176758,
+      "learning_rate": 2.016016016016016e-06,
+      "loss": 0.3295,
+      "step": 79760
+    },
+    {
+      "epoch": 239.55,
+      "grad_norm": 17.53756332397461,
+      "learning_rate": 2.015015015015015e-06,
+      "loss": 0.366,
+      "step": 79770
+    },
+    {
+      "epoch": 239.58,
+      "grad_norm": 18.31893539428711,
+      "learning_rate": 2.014014014014014e-06,
+      "loss": 0.3145,
+      "step": 79780
+    },
+    {
+      "epoch": 239.61,
+      "grad_norm": 17.69843101501465,
+      "learning_rate": 2.013013013013013e-06,
+      "loss": 0.3643,
+      "step": 79790
+    },
+    {
+      "epoch": 239.64,
+      "grad_norm": 17.619043350219727,
+      "learning_rate": 2.012012012012012e-06,
+      "loss": 0.3356,
+      "step": 79800
+    },
+    {
+      "epoch": 239.67,
+      "grad_norm": 10.754522323608398,
+      "learning_rate": 2.011011011011011e-06,
+      "loss": 0.2797,
+      "step": 79810
+    },
+    {
+      "epoch": 239.7,
+      "grad_norm": 17.77174186706543,
+      "learning_rate": 2.01001001001001e-06,
+      "loss": 0.3395,
+      "step": 79820
+    },
+    {
+      "epoch": 239.73,
+      "grad_norm": 18.076648712158203,
+      "learning_rate": 2.009009009009009e-06,
+      "loss": 0.3389,
+      "step": 79830
+    },
+    {
+      "epoch": 239.76,
+      "grad_norm": 15.888705253601074,
+      "learning_rate": 2.008008008008008e-06,
+      "loss": 0.3395,
+      "step": 79840
+    },
+    {
+      "epoch": 239.79,
+      "grad_norm": 23.647947311401367,
+      "learning_rate": 2.007007007007007e-06,
+      "loss": 0.3058,
+      "step": 79850
+    },
+    {
+      "epoch": 239.82,
+      "grad_norm": 13.498983383178711,
+      "learning_rate": 2.0060060060060062e-06,
+      "loss": 0.3074,
+      "step": 79860
+    },
+    {
+      "epoch": 239.85,
+      "grad_norm": 21.48940658569336,
+      "learning_rate": 2.0050050050050052e-06,
+      "loss": 0.349,
+      "step": 79870
+    },
+    {
+      "epoch": 239.88,
+      "grad_norm": 16.733036041259766,
+      "learning_rate": 2.0040040040040043e-06,
+      "loss": 0.3056,
+      "step": 79880
+    },
+    {
+      "epoch": 239.91,
+      "grad_norm": 10.600842475891113,
+      "learning_rate": 2.0030030030030033e-06,
+      "loss": 0.358,
+      "step": 79890
+    },
+    {
+      "epoch": 239.94,
+      "grad_norm": 9.175358772277832,
+      "learning_rate": 2.0020020020020023e-06,
+      "loss": 0.2988,
+      "step": 79900
+    },
+    {
+      "epoch": 239.97,
+      "grad_norm": 19.976301193237305,
+      "learning_rate": 2.0010010010010013e-06,
+      "loss": 0.3638,
+      "step": 79910
+    },
+    {
+      "epoch": 240.0,
+      "grad_norm": 11.870626449584961,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.2977,
+      "step": 79920
+    },
+    {
+      "epoch": 240.0,
+      "eval_accuracy": 0.8651,
+      "eval_loss": 0.5476164817810059,
+      "eval_runtime": 12.7316,
+      "eval_samples_per_second": 785.447,
+      "eval_steps_per_second": 3.142,
+      "step": 79920
+    },
+    {
+      "epoch": 240.03,
+      "grad_norm": 11.204678535461426,
+      "learning_rate": 1.998998998998999e-06,
+      "loss": 0.3033,
+      "step": 79930
+    },
+    {
+      "epoch": 240.06,
+      "grad_norm": 16.80437469482422,
+      "learning_rate": 1.9979979979979984e-06,
+      "loss": 0.3482,
+      "step": 79940
+    },
+    {
+      "epoch": 240.09,
+      "grad_norm": 28.398916244506836,
+      "learning_rate": 1.996996996996997e-06,
+      "loss": 0.3255,
+      "step": 79950
+    },
+    {
+      "epoch": 240.12,
+      "grad_norm": 21.19424057006836,
+      "learning_rate": 1.9959959959959964e-06,
+      "loss": 0.3676,
+      "step": 79960
+    },
+    {
+      "epoch": 240.15,
+      "grad_norm": 17.541101455688477,
+      "learning_rate": 1.994994994994995e-06,
+      "loss": 0.3689,
+      "step": 79970
+    },
+    {
+      "epoch": 240.18,
+      "grad_norm": 10.918506622314453,
+      "learning_rate": 1.9939939939939944e-06,
+      "loss": 0.3395,
+      "step": 79980
+    },
+    {
+      "epoch": 240.21,
+      "grad_norm": 16.14336395263672,
+      "learning_rate": 1.992992992992993e-06,
+      "loss": 0.3156,
+      "step": 79990
+    },
+    {
+      "epoch": 240.24,
+      "grad_norm": 15.821463584899902,
+      "learning_rate": 1.991991991991992e-06,
+      "loss": 0.2835,
+      "step": 80000
+    },
+    {
+      "epoch": 240.27,
+      "grad_norm": 17.794174194335938,
+      "learning_rate": 1.990990990990991e-06,
+      "loss": 0.3341,
+      "step": 80010
+    },
+    {
+      "epoch": 240.3,
+      "grad_norm": 11.795170783996582,
+      "learning_rate": 1.98998998998999e-06,
+      "loss": 0.3309,
+      "step": 80020
+    },
+    {
+      "epoch": 240.33,
+      "grad_norm": 16.69680404663086,
+      "learning_rate": 1.988988988988989e-06,
+      "loss": 0.3208,
+      "step": 80030
+    },
+    {
+      "epoch": 240.36,
+      "grad_norm": 17.55714225769043,
+      "learning_rate": 1.987987987987988e-06,
+      "loss": 0.2907,
+      "step": 80040
+    },
+    {
+      "epoch": 240.39,
+      "grad_norm": 16.08384132385254,
+      "learning_rate": 1.986986986986987e-06,
+      "loss": 0.3059,
+      "step": 80050
+    },
+    {
+      "epoch": 240.42,
+      "grad_norm": 10.468695640563965,
+      "learning_rate": 1.985985985985986e-06,
+      "loss": 0.3207,
+      "step": 80060
+    },
+    {
+      "epoch": 240.45,
+      "grad_norm": 18.90964126586914,
+      "learning_rate": 1.984984984984985e-06,
+      "loss": 0.3532,
+      "step": 80070
+    },
+    {
+      "epoch": 240.48,
+      "grad_norm": 22.214567184448242,
+      "learning_rate": 1.983983983983984e-06,
+      "loss": 0.3283,
+      "step": 80080
+    },
+    {
+      "epoch": 240.51,
+      "grad_norm": 15.861865997314453,
+      "learning_rate": 1.9829829829829832e-06,
+      "loss": 0.3244,
+      "step": 80090
+    },
+    {
+      "epoch": 240.54,
+      "grad_norm": 27.64719009399414,
+      "learning_rate": 1.9819819819819822e-06,
+      "loss": 0.3506,
+      "step": 80100
+    },
+    {
+      "epoch": 240.57,
+      "grad_norm": 14.036206245422363,
+      "learning_rate": 1.980980980980981e-06,
+      "loss": 0.3372,
+      "step": 80110
+    },
+    {
+      "epoch": 240.6,
+      "grad_norm": 16.551761627197266,
+      "learning_rate": 1.9799799799799803e-06,
+      "loss": 0.3183,
+      "step": 80120
+    },
+    {
+      "epoch": 240.63,
+      "grad_norm": 17.582090377807617,
+      "learning_rate": 1.978978978978979e-06,
+      "loss": 0.3294,
+      "step": 80130
+    },
+    {
+      "epoch": 240.66,
+      "grad_norm": 15.658019065856934,
+      "learning_rate": 1.9779779779779783e-06,
+      "loss": 0.3192,
+      "step": 80140
+    },
+    {
+      "epoch": 240.69,
+      "grad_norm": 14.663774490356445,
+      "learning_rate": 1.9769769769769773e-06,
+      "loss": 0.3401,
+      "step": 80150
+    },
+    {
+      "epoch": 240.72,
+      "grad_norm": 13.1045503616333,
+      "learning_rate": 1.975975975975976e-06,
+      "loss": 0.3889,
+      "step": 80160
+    },
+    {
+      "epoch": 240.75,
+      "grad_norm": 23.921010971069336,
+      "learning_rate": 1.9749749749749754e-06,
+      "loss": 0.3116,
+      "step": 80170
+    },
+    {
+      "epoch": 240.78,
+      "grad_norm": 17.094263076782227,
+      "learning_rate": 1.973973973973974e-06,
+      "loss": 0.3316,
+      "step": 80180
+    },
+    {
+      "epoch": 240.81,
+      "grad_norm": 11.58454418182373,
+      "learning_rate": 1.9729729729729734e-06,
+      "loss": 0.2848,
+      "step": 80190
+    },
+    {
+      "epoch": 240.84,
+      "grad_norm": 11.869529724121094,
+      "learning_rate": 1.971971971971972e-06,
+      "loss": 0.34,
+      "step": 80200
+    },
+    {
+      "epoch": 240.87,
+      "grad_norm": 15.963729858398438,
+      "learning_rate": 1.9709709709709714e-06,
+      "loss": 0.3112,
+      "step": 80210
+    },
+    {
+      "epoch": 240.9,
+      "grad_norm": 25.59613609313965,
+      "learning_rate": 1.96996996996997e-06,
+      "loss": 0.3339,
+      "step": 80220
+    },
+    {
+      "epoch": 240.93,
+      "grad_norm": 16.500625610351562,
+      "learning_rate": 1.968968968968969e-06,
+      "loss": 0.345,
+      "step": 80230
+    },
+    {
+      "epoch": 240.96,
+      "grad_norm": 18.28738784790039,
+      "learning_rate": 1.967967967967968e-06,
+      "loss": 0.3041,
+      "step": 80240
+    },
+    {
+      "epoch": 240.99,
+      "grad_norm": 12.8403959274292,
+      "learning_rate": 1.966966966966967e-06,
+      "loss": 0.3695,
+      "step": 80250
+    },
+    {
+      "epoch": 241.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5411624908447266,
+      "eval_runtime": 12.9212,
+      "eval_samples_per_second": 773.921,
+      "eval_steps_per_second": 3.096,
+      "step": 80253
+    },
+    {
+      "epoch": 241.02,
+      "grad_norm": 18.457366943359375,
+      "learning_rate": 1.965965965965966e-06,
+      "loss": 0.3011,
+      "step": 80260
+    },
+    {
+      "epoch": 241.05,
+      "grad_norm": 12.979890823364258,
+      "learning_rate": 1.964964964964965e-06,
+      "loss": 0.3317,
+      "step": 80270
+    },
+    {
+      "epoch": 241.08,
+      "grad_norm": 15.250653266906738,
+      "learning_rate": 1.963963963963964e-06,
+      "loss": 0.3466,
+      "step": 80280
+    },
+    {
+      "epoch": 241.11,
+      "grad_norm": 16.097787857055664,
+      "learning_rate": 1.962962962962963e-06,
+      "loss": 0.335,
+      "step": 80290
+    },
+    {
+      "epoch": 241.14,
+      "grad_norm": 21.92961311340332,
+      "learning_rate": 1.961961961961962e-06,
+      "loss": 0.3214,
+      "step": 80300
+    },
+    {
+      "epoch": 241.17,
+      "grad_norm": 11.578932762145996,
+      "learning_rate": 1.960960960960961e-06,
+      "loss": 0.3108,
+      "step": 80310
+    },
+    {
+      "epoch": 241.2,
+      "grad_norm": 15.344192504882812,
+      "learning_rate": 1.9599599599599602e-06,
+      "loss": 0.3051,
+      "step": 80320
+    },
+    {
+      "epoch": 241.23,
+      "grad_norm": 12.488163948059082,
+      "learning_rate": 1.9589589589589592e-06,
+      "loss": 0.2983,
+      "step": 80330
+    },
+    {
+      "epoch": 241.26,
+      "grad_norm": 16.774211883544922,
+      "learning_rate": 1.957957957957958e-06,
+      "loss": 0.378,
+      "step": 80340
+    },
+    {
+      "epoch": 241.29,
+      "grad_norm": 14.601218223571777,
+      "learning_rate": 1.9569569569569573e-06,
+      "loss": 0.3669,
+      "step": 80350
+    },
+    {
+      "epoch": 241.32,
+      "grad_norm": 13.272588729858398,
+      "learning_rate": 1.955955955955956e-06,
+      "loss": 0.2931,
+      "step": 80360
+    },
+    {
+      "epoch": 241.35,
+      "grad_norm": 13.028489112854004,
+      "learning_rate": 1.9549549549549553e-06,
+      "loss": 0.3295,
+      "step": 80370
+    },
+    {
+      "epoch": 241.38,
+      "grad_norm": 22.201093673706055,
+      "learning_rate": 1.953953953953954e-06,
+      "loss": 0.3447,
+      "step": 80380
+    },
+    {
+      "epoch": 241.41,
+      "grad_norm": 14.635054588317871,
+      "learning_rate": 1.952952952952953e-06,
+      "loss": 0.3305,
+      "step": 80390
+    },
+    {
+      "epoch": 241.44,
+      "grad_norm": 12.525761604309082,
+      "learning_rate": 1.951951951951952e-06,
+      "loss": 0.297,
+      "step": 80400
+    },
+    {
+      "epoch": 241.47,
+      "grad_norm": 14.617831230163574,
+      "learning_rate": 1.950950950950951e-06,
+      "loss": 0.3441,
+      "step": 80410
+    },
+    {
+      "epoch": 241.5,
+      "grad_norm": 14.27873706817627,
+      "learning_rate": 1.9499499499499504e-06,
+      "loss": 0.3259,
+      "step": 80420
+    },
+    {
+      "epoch": 241.53,
+      "grad_norm": 25.397733688354492,
+      "learning_rate": 1.948948948948949e-06,
+      "loss": 0.3477,
+      "step": 80430
+    },
+    {
+      "epoch": 241.56,
+      "grad_norm": 12.720926284790039,
+      "learning_rate": 1.947947947947948e-06,
+      "loss": 0.3238,
+      "step": 80440
+    },
+    {
+      "epoch": 241.59,
+      "grad_norm": 13.693068504333496,
+      "learning_rate": 1.946946946946947e-06,
+      "loss": 0.3441,
+      "step": 80450
+    },
+    {
+      "epoch": 241.62,
+      "grad_norm": 20.14032745361328,
+      "learning_rate": 1.945945945945946e-06,
+      "loss": 0.3233,
+      "step": 80460
+    },
+    {
+      "epoch": 241.65,
+      "grad_norm": 16.049837112426758,
+      "learning_rate": 1.944944944944945e-06,
+      "loss": 0.2958,
+      "step": 80470
+    },
+    {
+      "epoch": 241.68,
+      "grad_norm": 14.908256530761719,
+      "learning_rate": 1.943943943943944e-06,
+      "loss": 0.3269,
+      "step": 80480
+    },
+    {
+      "epoch": 241.71,
+      "grad_norm": 14.607721328735352,
+      "learning_rate": 1.942942942942943e-06,
+      "loss": 0.3315,
+      "step": 80490
+    },
+    {
+      "epoch": 241.74,
+      "grad_norm": 20.029481887817383,
+      "learning_rate": 1.941941941941942e-06,
+      "loss": 0.3562,
+      "step": 80500
+    },
+    {
+      "epoch": 241.77,
+      "grad_norm": 17.23569107055664,
+      "learning_rate": 1.940940940940941e-06,
+      "loss": 0.3359,
+      "step": 80510
+    },
+    {
+      "epoch": 241.8,
+      "grad_norm": 13.183422088623047,
+      "learning_rate": 1.93993993993994e-06,
+      "loss": 0.3167,
+      "step": 80520
+    },
+    {
+      "epoch": 241.83,
+      "grad_norm": 27.88653564453125,
+      "learning_rate": 1.938938938938939e-06,
+      "loss": 0.3308,
+      "step": 80530
+    },
+    {
+      "epoch": 241.86,
+      "grad_norm": 14.324834823608398,
+      "learning_rate": 1.937937937937938e-06,
+      "loss": 0.36,
+      "step": 80540
+    },
+    {
+      "epoch": 241.89,
+      "grad_norm": 12.241729736328125,
+      "learning_rate": 1.9369369369369372e-06,
+      "loss": 0.3171,
+      "step": 80550
+    },
+    {
+      "epoch": 241.92,
+      "grad_norm": 22.98464584350586,
+      "learning_rate": 1.9359359359359362e-06,
+      "loss": 0.3191,
+      "step": 80560
+    },
+    {
+      "epoch": 241.95,
+      "grad_norm": 15.394516944885254,
+      "learning_rate": 1.934934934934935e-06,
+      "loss": 0.3381,
+      "step": 80570
+    },
+    {
+      "epoch": 241.98,
+      "grad_norm": 12.9799222946167,
+      "learning_rate": 1.9339339339339343e-06,
+      "loss": 0.3683,
+      "step": 80580
+    },
+    {
+      "epoch": 242.0,
+      "eval_accuracy": 0.865,
+      "eval_loss": 0.5449464321136475,
+      "eval_runtime": 12.8157,
+      "eval_samples_per_second": 780.291,
+      "eval_steps_per_second": 3.121,
+      "step": 80586
+    },
+    {
+      "epoch": 242.01,
+      "grad_norm": 21.96697235107422,
+      "learning_rate": 1.932932932932933e-06,
+      "loss": 0.3387,
+      "step": 80590
+    },
+    {
+      "epoch": 242.04,
+      "grad_norm": 12.99114990234375,
+      "learning_rate": 1.9319319319319323e-06,
+      "loss": 0.299,
+      "step": 80600
+    },
+    {
+      "epoch": 242.07,
+      "grad_norm": 14.51639461517334,
+      "learning_rate": 1.930930930930931e-06,
+      "loss": 0.33,
+      "step": 80610
+    },
+    {
+      "epoch": 242.1,
+      "grad_norm": 14.608600616455078,
+      "learning_rate": 1.92992992992993e-06,
+      "loss": 0.3541,
+      "step": 80620
+    },
+    {
+      "epoch": 242.13,
+      "grad_norm": 10.987106323242188,
+      "learning_rate": 1.928928928928929e-06,
+      "loss": 0.3633,
+      "step": 80630
+    },
+    {
+      "epoch": 242.16,
+      "grad_norm": 14.02862548828125,
+      "learning_rate": 1.927927927927928e-06,
+      "loss": 0.3025,
+      "step": 80640
+    },
+    {
+      "epoch": 242.19,
+      "grad_norm": 8.949435234069824,
+      "learning_rate": 1.926926926926927e-06,
+      "loss": 0.3082,
+      "step": 80650
+    },
+    {
+      "epoch": 242.22,
+      "grad_norm": 17.92422866821289,
+      "learning_rate": 1.925925925925926e-06,
+      "loss": 0.3549,
+      "step": 80660
+    },
+    {
+      "epoch": 242.25,
+      "grad_norm": 14.74721908569336,
+      "learning_rate": 1.924924924924925e-06,
+      "loss": 0.3151,
+      "step": 80670
+    },
+    {
+      "epoch": 242.28,
+      "grad_norm": 14.593160629272461,
+      "learning_rate": 1.923923923923924e-06,
+      "loss": 0.3546,
+      "step": 80680
+    },
+    {
+      "epoch": 242.31,
+      "grad_norm": 20.33402442932129,
+      "learning_rate": 1.922922922922923e-06,
+      "loss": 0.3468,
+      "step": 80690
+    },
+    {
+      "epoch": 242.34,
+      "grad_norm": 16.579395294189453,
+      "learning_rate": 1.921921921921922e-06,
+      "loss": 0.3504,
+      "step": 80700
+    },
+    {
+      "epoch": 242.37,
+      "grad_norm": 20.77059555053711,
+      "learning_rate": 1.920920920920921e-06,
+      "loss": 0.3755,
+      "step": 80710
+    },
+    {
+      "epoch": 242.4,
+      "grad_norm": 15.104034423828125,
+      "learning_rate": 1.91991991991992e-06,
+      "loss": 0.3249,
+      "step": 80720
+    },
+    {
+      "epoch": 242.43,
+      "grad_norm": 15.134157180786133,
+      "learning_rate": 1.918918918918919e-06,
+      "loss": 0.372,
+      "step": 80730
+    },
+    {
+      "epoch": 242.46,
+      "grad_norm": 12.07043743133545,
+      "learning_rate": 1.917917917917918e-06,
+      "loss": 0.3463,
+      "step": 80740
+    },
+    {
+      "epoch": 242.49,
+      "grad_norm": 8.984689712524414,
+      "learning_rate": 1.916916916916917e-06,
+      "loss": 0.3193,
+      "step": 80750
+    },
+    {
+      "epoch": 242.52,
+      "grad_norm": 11.523752212524414,
+      "learning_rate": 1.915915915915916e-06,
+      "loss": 0.3526,
+      "step": 80760
+    },
+    {
+      "epoch": 242.55,
+      "grad_norm": 18.332786560058594,
+      "learning_rate": 1.914914914914915e-06,
+      "loss": 0.3109,
+      "step": 80770
+    },
+    {
+      "epoch": 242.58,
+      "grad_norm": 15.757955551147461,
+      "learning_rate": 1.9139139139139142e-06,
+      "loss": 0.3424,
+      "step": 80780
+    },
+    {
+      "epoch": 242.61,
+      "grad_norm": 18.288837432861328,
+      "learning_rate": 1.9129129129129132e-06,
+      "loss": 0.2901,
+      "step": 80790
+    },
+    {
+      "epoch": 242.64,
+      "grad_norm": 20.417924880981445,
+      "learning_rate": 1.911911911911912e-06,
+      "loss": 0.3905,
+      "step": 80800
+    },
+    {
+      "epoch": 242.67,
+      "grad_norm": 13.702827453613281,
+      "learning_rate": 1.9109109109109113e-06,
+      "loss": 0.3346,
+      "step": 80810
+    },
+    {
+      "epoch": 242.7,
+      "grad_norm": 15.495708465576172,
+      "learning_rate": 1.90990990990991e-06,
+      "loss": 0.3054,
+      "step": 80820
+    },
+    {
+      "epoch": 242.73,
+      "grad_norm": 17.911705017089844,
+      "learning_rate": 1.9089089089089093e-06,
+      "loss": 0.3336,
+      "step": 80830
+    },
+    {
+      "epoch": 242.76,
+      "grad_norm": 14.039260864257812,
+      "learning_rate": 1.907907907907908e-06,
+      "loss": 0.3054,
+      "step": 80840
+    },
+    {
+      "epoch": 242.79,
+      "grad_norm": 17.092914581298828,
+      "learning_rate": 1.9069069069069071e-06,
+      "loss": 0.3907,
+      "step": 80850
+    },
+    {
+      "epoch": 242.82,
+      "grad_norm": 13.11850643157959,
+      "learning_rate": 1.905905905905906e-06,
+      "loss": 0.3289,
+      "step": 80860
+    },
+    {
+      "epoch": 242.85,
+      "grad_norm": 18.23267364501953,
+      "learning_rate": 1.9049049049049052e-06,
+      "loss": 0.3423,
+      "step": 80870
+    },
+    {
+      "epoch": 242.88,
+      "grad_norm": 15.686576843261719,
+      "learning_rate": 1.903903903903904e-06,
+      "loss": 0.2882,
+      "step": 80880
+    },
+    {
+      "epoch": 242.91,
+      "grad_norm": 25.537336349487305,
+      "learning_rate": 1.902902902902903e-06,
+      "loss": 0.3127,
+      "step": 80890
+    },
+    {
+      "epoch": 242.94,
+      "grad_norm": 16.806730270385742,
+      "learning_rate": 1.901901901901902e-06,
+      "loss": 0.3688,
+      "step": 80900
+    },
+    {
+      "epoch": 242.97,
+      "grad_norm": 17.412036895751953,
+      "learning_rate": 1.900900900900901e-06,
+      "loss": 0.2971,
+      "step": 80910
+    },
+    {
+      "epoch": 243.0,
+      "eval_accuracy": 0.8658,
+      "eval_loss": 0.548987090587616,
+      "eval_runtime": 12.999,
+      "eval_samples_per_second": 769.288,
+      "eval_steps_per_second": 3.077,
+      "step": 80919
+    },
+    {
+      "epoch": 243.0,
+      "grad_norm": 21.523405075073242,
+      "learning_rate": 1.8998998998998998e-06,
+      "loss": 0.3186,
+      "step": 80920
+    },
+    {
+      "epoch": 243.03,
+      "grad_norm": 12.608552932739258,
+      "learning_rate": 1.898898898898899e-06,
+      "loss": 0.4119,
+      "step": 80930
+    },
+    {
+      "epoch": 243.06,
+      "grad_norm": 17.777620315551758,
+      "learning_rate": 1.897897897897898e-06,
+      "loss": 0.3383,
+      "step": 80940
+    },
+    {
+      "epoch": 243.09,
+      "grad_norm": 15.450210571289062,
+      "learning_rate": 1.896896896896897e-06,
+      "loss": 0.3261,
+      "step": 80950
+    },
+    {
+      "epoch": 243.12,
+      "grad_norm": 17.033166885375977,
+      "learning_rate": 1.8958958958958961e-06,
+      "loss": 0.3488,
+      "step": 80960
+    },
+    {
+      "epoch": 243.15,
+      "grad_norm": 20.467172622680664,
+      "learning_rate": 1.894894894894895e-06,
+      "loss": 0.37,
+      "step": 80970
+    },
+    {
+      "epoch": 243.18,
+      "grad_norm": 25.753999710083008,
+      "learning_rate": 1.8938938938938942e-06,
+      "loss": 0.3647,
+      "step": 80980
+    },
+    {
+      "epoch": 243.21,
+      "grad_norm": 24.257617950439453,
+      "learning_rate": 1.892892892892893e-06,
+      "loss": 0.312,
+      "step": 80990
+    },
+    {
+      "epoch": 243.24,
+      "grad_norm": 10.119921684265137,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.3201,
+      "step": 81000
+    },
+    {
+      "epoch": 243.27,
+      "grad_norm": 17.406879425048828,
+      "learning_rate": 1.890890890890891e-06,
+      "loss": 0.3724,
+      "step": 81010
+    },
+    {
+      "epoch": 243.3,
+      "grad_norm": 17.87641143798828,
+      "learning_rate": 1.8898898898898902e-06,
+      "loss": 0.3196,
+      "step": 81020
+    },
+    {
+      "epoch": 243.33,
+      "grad_norm": 12.264856338500977,
+      "learning_rate": 1.888888888888889e-06,
+      "loss": 0.3151,
+      "step": 81030
+    },
+    {
+      "epoch": 243.36,
+      "grad_norm": 15.00865650177002,
+      "learning_rate": 1.887887887887888e-06,
+      "loss": 0.3506,
+      "step": 81040
+    },
+    {
+      "epoch": 243.39,
+      "grad_norm": 11.05329418182373,
+      "learning_rate": 1.886886886886887e-06,
+      "loss": 0.3179,
+      "step": 81050
+    },
+    {
+      "epoch": 243.42,
+      "grad_norm": 15.361865043640137,
+      "learning_rate": 1.885885885885886e-06,
+      "loss": 0.3393,
+      "step": 81060
+    },
+    {
+      "epoch": 243.45,
+      "grad_norm": 14.141924858093262,
+      "learning_rate": 1.884884884884885e-06,
+      "loss": 0.3251,
+      "step": 81070
+    },
+    {
+      "epoch": 243.48,
+      "grad_norm": 22.012771606445312,
+      "learning_rate": 1.8838838838838841e-06,
+      "loss": 0.3623,
+      "step": 81080
+    },
+    {
+      "epoch": 243.51,
+      "grad_norm": 19.159448623657227,
+      "learning_rate": 1.882882882882883e-06,
+      "loss": 0.3516,
+      "step": 81090
+    },
+    {
+      "epoch": 243.54,
+      "grad_norm": 22.87691307067871,
+      "learning_rate": 1.8818818818818822e-06,
+      "loss": 0.3481,
+      "step": 81100
+    },
+    {
+      "epoch": 243.57,
+      "grad_norm": 10.513298034667969,
+      "learning_rate": 1.880880880880881e-06,
+      "loss": 0.3393,
+      "step": 81110
+    },
+    {
+      "epoch": 243.6,
+      "grad_norm": 16.58595848083496,
+      "learning_rate": 1.87987987987988e-06,
+      "loss": 0.2873,
+      "step": 81120
+    },
+    {
+      "epoch": 243.63,
+      "grad_norm": 22.296300888061523,
+      "learning_rate": 1.878878878878879e-06,
+      "loss": 0.3648,
+      "step": 81130
+    },
+    {
+      "epoch": 243.66,
+      "grad_norm": 13.962804794311523,
+      "learning_rate": 1.877877877877878e-06,
+      "loss": 0.2791,
+      "step": 81140
+    },
+    {
+      "epoch": 243.69,
+      "grad_norm": 12.434715270996094,
+      "learning_rate": 1.8768768768768768e-06,
+      "loss": 0.3295,
+      "step": 81150
+    },
+    {
+      "epoch": 243.72,
+      "grad_norm": 13.211495399475098,
+      "learning_rate": 1.875875875875876e-06,
+      "loss": 0.3755,
+      "step": 81160
+    },
+    {
+      "epoch": 243.75,
+      "grad_norm": 18.29191017150879,
+      "learning_rate": 1.8748748748748749e-06,
+      "loss": 0.3826,
+      "step": 81170
+    },
+    {
+      "epoch": 243.78,
+      "grad_norm": 20.85376739501953,
+      "learning_rate": 1.873873873873874e-06,
+      "loss": 0.3506,
+      "step": 81180
+    },
+    {
+      "epoch": 243.81,
+      "grad_norm": 9.966811180114746,
+      "learning_rate": 1.872872872872873e-06,
+      "loss": 0.3331,
+      "step": 81190
+    },
+    {
+      "epoch": 243.84,
+      "grad_norm": 18.05333137512207,
+      "learning_rate": 1.871871871871872e-06,
+      "loss": 0.3053,
+      "step": 81200
+    },
+    {
+      "epoch": 243.87,
+      "grad_norm": 9.762744903564453,
+      "learning_rate": 1.8708708708708712e-06,
+      "loss": 0.333,
+      "step": 81210
+    },
+    {
+      "epoch": 243.9,
+      "grad_norm": 16.034713745117188,
+      "learning_rate": 1.86986986986987e-06,
+      "loss": 0.3621,
+      "step": 81220
+    },
+    {
+      "epoch": 243.93,
+      "grad_norm": 17.270957946777344,
+      "learning_rate": 1.8688688688688692e-06,
+      "loss": 0.3732,
+      "step": 81230
+    },
+    {
+      "epoch": 243.96,
+      "grad_norm": 14.207944869995117,
+      "learning_rate": 1.867867867867868e-06,
+      "loss": 0.3294,
+      "step": 81240
+    },
+    {
+      "epoch": 243.99,
+      "grad_norm": 10.896373748779297,
+      "learning_rate": 1.8668668668668672e-06,
+      "loss": 0.3532,
+      "step": 81250
+    },
+    {
+      "epoch": 244.0,
+      "eval_accuracy": 0.8666,
+      "eval_loss": 0.5449116826057434,
+      "eval_runtime": 12.5696,
+      "eval_samples_per_second": 795.57,
+      "eval_steps_per_second": 3.182,
+      "step": 81252
+    },
+    {
+      "epoch": 244.02,
+      "grad_norm": 19.189836502075195,
+      "learning_rate": 1.865865865865866e-06,
+      "loss": 0.3331,
+      "step": 81260
+    },
+    {
+      "epoch": 244.05,
+      "grad_norm": 12.436052322387695,
+      "learning_rate": 1.864864864864865e-06,
+      "loss": 0.3419,
+      "step": 81270
+    },
+    {
+      "epoch": 244.08,
+      "grad_norm": 30.980087280273438,
+      "learning_rate": 1.863863863863864e-06,
+      "loss": 0.2992,
+      "step": 81280
+    },
+    {
+      "epoch": 244.11,
+      "grad_norm": 13.817534446716309,
+      "learning_rate": 1.862862862862863e-06,
+      "loss": 0.3138,
+      "step": 81290
+    },
+    {
+      "epoch": 244.14,
+      "grad_norm": 19.90482521057129,
+      "learning_rate": 1.861861861861862e-06,
+      "loss": 0.3068,
+      "step": 81300
+    },
+    {
+      "epoch": 244.17,
+      "grad_norm": 15.70452880859375,
+      "learning_rate": 1.8608608608608611e-06,
+      "loss": 0.3304,
+      "step": 81310
+    },
+    {
+      "epoch": 244.2,
+      "grad_norm": 13.34378719329834,
+      "learning_rate": 1.85985985985986e-06,
+      "loss": 0.3563,
+      "step": 81320
+    },
+    {
+      "epoch": 244.23,
+      "grad_norm": 14.720455169677734,
+      "learning_rate": 1.8588588588588592e-06,
+      "loss": 0.3248,
+      "step": 81330
+    },
+    {
+      "epoch": 244.26,
+      "grad_norm": 13.364724159240723,
+      "learning_rate": 1.857857857857858e-06,
+      "loss": 0.3111,
+      "step": 81340
+    },
+    {
+      "epoch": 244.29,
+      "grad_norm": 12.578872680664062,
+      "learning_rate": 1.856856856856857e-06,
+      "loss": 0.3382,
+      "step": 81350
+    },
+    {
+      "epoch": 244.32,
+      "grad_norm": 17.479238510131836,
+      "learning_rate": 1.855855855855856e-06,
+      "loss": 0.3396,
+      "step": 81360
+    },
+    {
+      "epoch": 244.35,
+      "grad_norm": 16.991031646728516,
+      "learning_rate": 1.854854854854855e-06,
+      "loss": 0.326,
+      "step": 81370
+    },
+    {
+      "epoch": 244.38,
+      "grad_norm": 12.88371753692627,
+      "learning_rate": 1.8538538538538538e-06,
+      "loss": 0.343,
+      "step": 81380
+    },
+    {
+      "epoch": 244.41,
+      "grad_norm": 17.05462646484375,
+      "learning_rate": 1.852852852852853e-06,
+      "loss": 0.3941,
+      "step": 81390
+    },
+    {
+      "epoch": 244.44,
+      "grad_norm": 14.899702072143555,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.366,
+      "step": 81400
+    },
+    {
+      "epoch": 244.47,
+      "grad_norm": 15.974350929260254,
+      "learning_rate": 1.850850850850851e-06,
+      "loss": 0.2942,
+      "step": 81410
+    },
+    {
+      "epoch": 244.5,
+      "grad_norm": 16.89360237121582,
+      "learning_rate": 1.84984984984985e-06,
+      "loss": 0.3182,
+      "step": 81420
+    },
+    {
+      "epoch": 244.53,
+      "grad_norm": 13.895329475402832,
+      "learning_rate": 1.848848848848849e-06,
+      "loss": 0.3002,
+      "step": 81430
+    },
+    {
+      "epoch": 244.56,
+      "grad_norm": 15.853198051452637,
+      "learning_rate": 1.847847847847848e-06,
+      "loss": 0.3047,
+      "step": 81440
+    },
+    {
+      "epoch": 244.59,
+      "grad_norm": 20.55255126953125,
+      "learning_rate": 1.846846846846847e-06,
+      "loss": 0.3296,
+      "step": 81450
+    },
+    {
+      "epoch": 244.62,
+      "grad_norm": 10.173460006713867,
+      "learning_rate": 1.8458458458458458e-06,
+      "loss": 0.3145,
+      "step": 81460
+    },
+    {
+      "epoch": 244.65,
+      "grad_norm": 22.941814422607422,
+      "learning_rate": 1.844844844844845e-06,
+      "loss": 0.3513,
+      "step": 81470
+    },
+    {
+      "epoch": 244.68,
+      "grad_norm": 16.417139053344727,
+      "learning_rate": 1.8438438438438442e-06,
+      "loss": 0.3568,
+      "step": 81480
+    },
+    {
+      "epoch": 244.71,
+      "grad_norm": 22.05870246887207,
+      "learning_rate": 1.842842842842843e-06,
+      "loss": 0.3223,
+      "step": 81490
+    },
+    {
+      "epoch": 244.74,
+      "grad_norm": 12.242451667785645,
+      "learning_rate": 1.841841841841842e-06,
+      "loss": 0.4026,
+      "step": 81500
+    },
+    {
+      "epoch": 244.77,
+      "grad_norm": 24.37861442565918,
+      "learning_rate": 1.840840840840841e-06,
+      "loss": 0.344,
+      "step": 81510
+    },
+    {
+      "epoch": 244.8,
+      "grad_norm": 17.88088035583496,
+      "learning_rate": 1.83983983983984e-06,
+      "loss": 0.3386,
+      "step": 81520
+    },
+    {
+      "epoch": 244.83,
+      "grad_norm": 21.318689346313477,
+      "learning_rate": 1.838838838838839e-06,
+      "loss": 0.3208,
+      "step": 81530
+    },
+    {
+      "epoch": 244.86,
+      "grad_norm": 19.77469825744629,
+      "learning_rate": 1.8378378378378381e-06,
+      "loss": 0.3312,
+      "step": 81540
+    },
+    {
+      "epoch": 244.89,
+      "grad_norm": 19.20047950744629,
+      "learning_rate": 1.836836836836837e-06,
+      "loss": 0.3179,
+      "step": 81550
+    },
+    {
+      "epoch": 244.92,
+      "grad_norm": 9.351749420166016,
+      "learning_rate": 1.8358358358358362e-06,
+      "loss": 0.292,
+      "step": 81560
+    },
+    {
+      "epoch": 244.95,
+      "grad_norm": 19.025081634521484,
+      "learning_rate": 1.834834834834835e-06,
+      "loss": 0.3345,
+      "step": 81570
+    },
+    {
+      "epoch": 244.98,
+      "grad_norm": 13.487537384033203,
+      "learning_rate": 1.833833833833834e-06,
+      "loss": 0.3014,
+      "step": 81580
+    },
+    {
+      "epoch": 245.0,
+      "eval_accuracy": 0.8657,
+      "eval_loss": 0.5447872877120972,
+      "eval_runtime": 12.7634,
+      "eval_samples_per_second": 783.49,
+      "eval_steps_per_second": 3.134,
+      "step": 81585
+    },
+    {
+      "epoch": 245.02,
+      "grad_norm": 19.5460147857666,
+      "learning_rate": 1.832832832832833e-06,
+      "loss": 0.3108,
+      "step": 81590
+    },
+    {
+      "epoch": 245.05,
+      "grad_norm": 13.650753021240234,
+      "learning_rate": 1.831831831831832e-06,
+      "loss": 0.3207,
+      "step": 81600
+    },
+    {
+      "epoch": 245.08,
+      "grad_norm": 9.89492130279541,
+      "learning_rate": 1.8308308308308308e-06,
+      "loss": 0.3158,
+      "step": 81610
+    },
+    {
+      "epoch": 245.11,
+      "grad_norm": 14.215547561645508,
+      "learning_rate": 1.82982982982983e-06,
+      "loss": 0.321,
+      "step": 81620
+    },
+    {
+      "epoch": 245.14,
+      "grad_norm": 13.665942192077637,
+      "learning_rate": 1.8288288288288289e-06,
+      "loss": 0.3415,
+      "step": 81630
+    },
+    {
+      "epoch": 245.17,
+      "grad_norm": 19.17864990234375,
+      "learning_rate": 1.827827827827828e-06,
+      "loss": 0.3131,
+      "step": 81640
+    },
+    {
+      "epoch": 245.2,
+      "grad_norm": 17.144140243530273,
+      "learning_rate": 1.826826826826827e-06,
+      "loss": 0.3263,
+      "step": 81650
+    },
+    {
+      "epoch": 245.23,
+      "grad_norm": 14.1176176071167,
+      "learning_rate": 1.825825825825826e-06,
+      "loss": 0.2863,
+      "step": 81660
+    },
+    {
+      "epoch": 245.26,
+      "grad_norm": 14.017192840576172,
+      "learning_rate": 1.824824824824825e-06,
+      "loss": 0.2998,
+      "step": 81670
+    },
+    {
+      "epoch": 245.29,
+      "grad_norm": 20.169992446899414,
+      "learning_rate": 1.823823823823824e-06,
+      "loss": 0.3158,
+      "step": 81680
+    },
+    {
+      "epoch": 245.32,
+      "grad_norm": 12.694307327270508,
+      "learning_rate": 1.8228228228228228e-06,
+      "loss": 0.3396,
+      "step": 81690
+    },
+    {
+      "epoch": 245.35,
+      "grad_norm": 23.84914207458496,
+      "learning_rate": 1.821821821821822e-06,
+      "loss": 0.361,
+      "step": 81700
+    },
+    {
+      "epoch": 245.38,
+      "grad_norm": 18.75893211364746,
+      "learning_rate": 1.8208208208208208e-06,
+      "loss": 0.421,
+      "step": 81710
+    },
+    {
+      "epoch": 245.41,
+      "grad_norm": 12.41589641571045,
+      "learning_rate": 1.81981981981982e-06,
+      "loss": 0.3715,
+      "step": 81720
+    },
+    {
+      "epoch": 245.44,
+      "grad_norm": 13.896773338317871,
+      "learning_rate": 1.8188188188188188e-06,
+      "loss": 0.3334,
+      "step": 81730
+    },
+    {
+      "epoch": 245.47,
+      "grad_norm": 14.560100555419922,
+      "learning_rate": 1.817817817817818e-06,
+      "loss": 0.3651,
+      "step": 81740
+    },
+    {
+      "epoch": 245.5,
+      "grad_norm": 18.863901138305664,
+      "learning_rate": 1.816816816816817e-06,
+      "loss": 0.269,
+      "step": 81750
+    },
+    {
+      "epoch": 245.53,
+      "grad_norm": 14.717557907104492,
+      "learning_rate": 1.815815815815816e-06,
+      "loss": 0.3502,
+      "step": 81760
+    },
+    {
+      "epoch": 245.56,
+      "grad_norm": 18.017122268676758,
+      "learning_rate": 1.8148148148148151e-06,
+      "loss": 0.3387,
+      "step": 81770
+    },
+    {
+      "epoch": 245.59,
+      "grad_norm": 15.83314037322998,
+      "learning_rate": 1.813813813813814e-06,
+      "loss": 0.319,
+      "step": 81780
+    },
+    {
+      "epoch": 245.62,
+      "grad_norm": 18.767669677734375,
+      "learning_rate": 1.8128128128128132e-06,
+      "loss": 0.3463,
+      "step": 81790
+    },
+    {
+      "epoch": 245.65,
+      "grad_norm": 17.159120559692383,
+      "learning_rate": 1.811811811811812e-06,
+      "loss": 0.3202,
+      "step": 81800
+    },
+    {
+      "epoch": 245.68,
+      "grad_norm": 8.282094955444336,
+      "learning_rate": 1.810810810810811e-06,
+      "loss": 0.2781,
+      "step": 81810
+    },
+    {
+      "epoch": 245.71,
+      "grad_norm": 16.597396850585938,
+      "learning_rate": 1.80980980980981e-06,
+      "loss": 0.3019,
+      "step": 81820
+    },
+    {
+      "epoch": 245.74,
+      "grad_norm": 23.584762573242188,
+      "learning_rate": 1.808808808808809e-06,
+      "loss": 0.3429,
+      "step": 81830
+    },
+    {
+      "epoch": 245.77,
+      "grad_norm": 16.35009002685547,
+      "learning_rate": 1.8078078078078078e-06,
+      "loss": 0.313,
+      "step": 81840
+    },
+    {
+      "epoch": 245.8,
+      "grad_norm": 16.940536499023438,
+      "learning_rate": 1.806806806806807e-06,
+      "loss": 0.3289,
+      "step": 81850
+    },
+    {
+      "epoch": 245.83,
+      "grad_norm": 12.512069702148438,
+      "learning_rate": 1.8058058058058059e-06,
+      "loss": 0.2992,
+      "step": 81860
+    },
+    {
+      "epoch": 245.86,
+      "grad_norm": 12.59357738494873,
+      "learning_rate": 1.804804804804805e-06,
+      "loss": 0.3523,
+      "step": 81870
+    },
+    {
+      "epoch": 245.89,
+      "grad_norm": 14.0105619430542,
+      "learning_rate": 1.803803803803804e-06,
+      "loss": 0.3247,
+      "step": 81880
+    },
+    {
+      "epoch": 245.92,
+      "grad_norm": 9.869412422180176,
+      "learning_rate": 1.802802802802803e-06,
+      "loss": 0.3328,
+      "step": 81890
+    },
+    {
+      "epoch": 245.95,
+      "grad_norm": 13.540960311889648,
+      "learning_rate": 1.801801801801802e-06,
+      "loss": 0.3813,
+      "step": 81900
+    },
+    {
+      "epoch": 245.98,
+      "grad_norm": 19.825885772705078,
+      "learning_rate": 1.800800800800801e-06,
+      "loss": 0.4048,
+      "step": 81910
+    },
+    {
+      "epoch": 246.0,
+      "eval_accuracy": 0.8654,
+      "eval_loss": 0.546809732913971,
+      "eval_runtime": 12.4941,
+      "eval_samples_per_second": 800.38,
+      "eval_steps_per_second": 3.202,
+      "step": 81918
+    },
+    {
+      "epoch": 246.01,
+      "grad_norm": 13.612279891967773,
+      "learning_rate": 1.7997997997997998e-06,
+      "loss": 0.3746,
+      "step": 81920
+    },
+    {
+      "epoch": 246.04,
+      "grad_norm": 14.892333030700684,
+      "learning_rate": 1.798798798798799e-06,
+      "loss": 0.2988,
+      "step": 81930
+    },
+    {
+      "epoch": 246.07,
+      "grad_norm": 13.488129615783691,
+      "learning_rate": 1.7977977977977978e-06,
+      "loss": 0.3487,
+      "step": 81940
+    },
+    {
+      "epoch": 246.1,
+      "grad_norm": 10.637248039245605,
+      "learning_rate": 1.796796796796797e-06,
+      "loss": 0.3192,
+      "step": 81950
+    },
+    {
+      "epoch": 246.13,
+      "grad_norm": 7.4920525550842285,
+      "learning_rate": 1.7957957957957958e-06,
+      "loss": 0.3327,
+      "step": 81960
+    },
+    {
+      "epoch": 246.16,
+      "grad_norm": 12.588641166687012,
+      "learning_rate": 1.7947947947947949e-06,
+      "loss": 0.3275,
+      "step": 81970
+    },
+    {
+      "epoch": 246.19,
+      "grad_norm": 10.419391632080078,
+      "learning_rate": 1.7937937937937939e-06,
+      "loss": 0.3323,
+      "step": 81980
+    },
+    {
+      "epoch": 246.22,
+      "grad_norm": 16.017789840698242,
+      "learning_rate": 1.7927927927927929e-06,
+      "loss": 0.3496,
+      "step": 81990
+    },
+    {
+      "epoch": 246.25,
+      "grad_norm": 15.92422866821289,
+      "learning_rate": 1.7917917917917917e-06,
+      "loss": 0.3203,
+      "step": 82000
+    },
+    {
+      "epoch": 246.28,
+      "grad_norm": 12.710223197937012,
+      "learning_rate": 1.790790790790791e-06,
+      "loss": 0.3203,
+      "step": 82010
+    },
+    {
+      "epoch": 246.31,
+      "grad_norm": 21.821687698364258,
+      "learning_rate": 1.7897897897897902e-06,
+      "loss": 0.3304,
+      "step": 82020
+    },
+    {
+      "epoch": 246.34,
+      "grad_norm": 17.396705627441406,
+      "learning_rate": 1.788788788788789e-06,
+      "loss": 0.3149,
+      "step": 82030
+    },
+    {
+      "epoch": 246.37,
+      "grad_norm": 9.572152137756348,
+      "learning_rate": 1.787787787787788e-06,
+      "loss": 0.2665,
+      "step": 82040
+    },
+    {
+      "epoch": 246.4,
+      "grad_norm": 15.071588516235352,
+      "learning_rate": 1.786786786786787e-06,
+      "loss": 0.365,
+      "step": 82050
+    },
+    {
+      "epoch": 246.43,
+      "grad_norm": 11.388834953308105,
+      "learning_rate": 1.785785785785786e-06,
+      "loss": 0.3056,
+      "step": 82060
+    },
+    {
+      "epoch": 246.46,
+      "grad_norm": 12.682565689086914,
+      "learning_rate": 1.7847847847847848e-06,
+      "loss": 0.3267,
+      "step": 82070
+    },
+    {
+      "epoch": 246.49,
+      "grad_norm": 14.248442649841309,
+      "learning_rate": 1.783783783783784e-06,
+      "loss": 0.3071,
+      "step": 82080
+    },
+    {
+      "epoch": 246.52,
+      "grad_norm": 17.240930557250977,
+      "learning_rate": 1.7827827827827829e-06,
+      "loss": 0.3541,
+      "step": 82090
+    },
+    {
+      "epoch": 246.55,
+      "grad_norm": 19.38409423828125,
+      "learning_rate": 1.781781781781782e-06,
+      "loss": 0.3079,
+      "step": 82100
+    },
+    {
+      "epoch": 246.58,
+      "grad_norm": 14.663848876953125,
+      "learning_rate": 1.780780780780781e-06,
+      "loss": 0.3111,
+      "step": 82110
+    },
+    {
+      "epoch": 246.61,
+      "grad_norm": 28.353849411010742,
+      "learning_rate": 1.77977977977978e-06,
+      "loss": 0.3283,
+      "step": 82120
+    },
+    {
+      "epoch": 246.64,
+      "grad_norm": 17.889808654785156,
+      "learning_rate": 1.778778778778779e-06,
+      "loss": 0.3413,
+      "step": 82130
+    },
+    {
+      "epoch": 246.67,
+      "grad_norm": 18.01947593688965,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 0.4079,
+      "step": 82140
+    },
+    {
+      "epoch": 246.7,
+      "grad_norm": 18.566553115844727,
+      "learning_rate": 1.7767767767767768e-06,
+      "loss": 0.3158,
+      "step": 82150
+    },
+    {
+      "epoch": 246.73,
+      "grad_norm": 21.085205078125,
+      "learning_rate": 1.775775775775776e-06,
+      "loss": 0.3422,
+      "step": 82160
+    },
+    {
+      "epoch": 246.76,
+      "grad_norm": 9.938655853271484,
+      "learning_rate": 1.7747747747747748e-06,
+      "loss": 0.295,
+      "step": 82170
+    },
+    {
+      "epoch": 246.79,
+      "grad_norm": 17.590131759643555,
+      "learning_rate": 1.773773773773774e-06,
+      "loss": 0.3467,
+      "step": 82180
+    },
+    {
+      "epoch": 246.82,
+      "grad_norm": 14.689308166503906,
+      "learning_rate": 1.7727727727727728e-06,
+      "loss": 0.3509,
+      "step": 82190
+    },
+    {
+      "epoch": 246.85,
+      "grad_norm": 17.561494827270508,
+      "learning_rate": 1.7717717717717719e-06,
+      "loss": 0.3427,
+      "step": 82200
+    },
+    {
+      "epoch": 246.88,
+      "grad_norm": 12.415604591369629,
+      "learning_rate": 1.7707707707707709e-06,
+      "loss": 0.3462,
+      "step": 82210
+    },
+    {
+      "epoch": 246.91,
+      "grad_norm": 17.204002380371094,
+      "learning_rate": 1.7697697697697699e-06,
+      "loss": 0.3288,
+      "step": 82220
+    },
+    {
+      "epoch": 246.94,
+      "grad_norm": 15.313666343688965,
+      "learning_rate": 1.7687687687687687e-06,
+      "loss": 0.3017,
+      "step": 82230
+    },
+    {
+      "epoch": 246.97,
+      "grad_norm": 17.271604537963867,
+      "learning_rate": 1.767767767767768e-06,
+      "loss": 0.3649,
+      "step": 82240
+    },
+    {
+      "epoch": 247.0,
+      "grad_norm": 12.725098609924316,
+      "learning_rate": 1.7667667667667667e-06,
+      "loss": 0.3279,
+      "step": 82250
+    },
+    {
+      "epoch": 247.0,
+      "eval_accuracy": 0.8661,
+      "eval_loss": 0.5501131415367126,
+      "eval_runtime": 12.6044,
+      "eval_samples_per_second": 793.371,
+      "eval_steps_per_second": 3.173,
+      "step": 82251
+    },
+    {
+      "epoch": 247.03,
+      "grad_norm": 16.746662139892578,
+      "learning_rate": 1.765765765765766e-06,
+      "loss": 0.567,
+      "step": 82260
+    },
+    {
+      "epoch": 247.06,
+      "grad_norm": 16.37392807006836,
+      "learning_rate": 1.7647647647647648e-06,
+      "loss": 0.359,
+      "step": 82270
+    },
+    {
+      "epoch": 247.09,
+      "grad_norm": 15.50261116027832,
+      "learning_rate": 1.763763763763764e-06,
+      "loss": 0.3502,
+      "step": 82280
+    },
+    {
+      "epoch": 247.12,
+      "grad_norm": 15.447113037109375,
+      "learning_rate": 1.762762762762763e-06,
+      "loss": 0.3307,
+      "step": 82290
+    },
+    {
+      "epoch": 247.15,
+      "grad_norm": 23.406131744384766,
+      "learning_rate": 1.7617617617617618e-06,
+      "loss": 0.3172,
+      "step": 82300
+    },
+    {
+      "epoch": 247.18,
+      "grad_norm": 12.724702835083008,
+      "learning_rate": 1.760760760760761e-06,
+      "loss": 0.323,
+      "step": 82310
+    },
+    {
+      "epoch": 247.21,
+      "grad_norm": 16.205799102783203,
+      "learning_rate": 1.7597597597597599e-06,
+      "loss": 0.333,
+      "step": 82320
+    },
+    {
+      "epoch": 247.24,
+      "grad_norm": 14.652810096740723,
+      "learning_rate": 1.758758758758759e-06,
+      "loss": 0.3159,
+      "step": 82330
+    },
+    {
+      "epoch": 247.27,
+      "grad_norm": 25.670320510864258,
+      "learning_rate": 1.757757757757758e-06,
+      "loss": 0.3472,
+      "step": 82340
+    },
+    {
+      "epoch": 247.3,
+      "grad_norm": 17.45107650756836,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.3001,
+      "step": 82350
+    },
+    {
+      "epoch": 247.33,
+      "grad_norm": 11.780815124511719,
+      "learning_rate": 1.755755755755756e-06,
+      "loss": 0.3125,
+      "step": 82360
+    },
+    {
+      "epoch": 247.36,
+      "grad_norm": 12.995439529418945,
+      "learning_rate": 1.754754754754755e-06,
+      "loss": 0.3119,
+      "step": 82370
+    },
+    {
+      "epoch": 247.39,
+      "grad_norm": 14.06275463104248,
+      "learning_rate": 1.7537537537537538e-06,
+      "loss": 0.3176,
+      "step": 82380
+    },
+    {
+      "epoch": 247.42,
+      "grad_norm": 8.032919883728027,
+      "learning_rate": 1.752752752752753e-06,
+      "loss": 0.2885,
+      "step": 82390
+    },
+    {
+      "epoch": 247.45,
+      "grad_norm": 33.870941162109375,
+      "learning_rate": 1.7517517517517518e-06,
+      "loss": 0.436,
+      "step": 82400
+    },
+    {
+      "epoch": 247.48,
+      "grad_norm": 17.057634353637695,
+      "learning_rate": 1.750750750750751e-06,
+      "loss": 0.3215,
+      "step": 82410
+    },
+    {
+      "epoch": 247.51,
+      "grad_norm": 21.164278030395508,
+      "learning_rate": 1.7497497497497498e-06,
+      "loss": 0.3306,
+      "step": 82420
+    },
+    {
+      "epoch": 247.54,
+      "grad_norm": 21.506759643554688,
+      "learning_rate": 1.7487487487487489e-06,
+      "loss": 0.3133,
+      "step": 82430
+    },
+    {
+      "epoch": 247.57,
+      "grad_norm": 13.116135597229004,
+      "learning_rate": 1.7477477477477479e-06,
+      "loss": 0.317,
+      "step": 82440
+    },
+    {
+      "epoch": 247.6,
+      "grad_norm": 11.89586353302002,
+      "learning_rate": 1.7467467467467469e-06,
+      "loss": 0.3634,
+      "step": 82450
+    },
+    {
+      "epoch": 247.63,
+      "grad_norm": 19.10785484313965,
+      "learning_rate": 1.7457457457457457e-06,
+      "loss": 0.368,
+      "step": 82460
+    },
+    {
+      "epoch": 247.66,
+      "grad_norm": 14.623037338256836,
+      "learning_rate": 1.744744744744745e-06,
+      "loss": 0.3092,
+      "step": 82470
+    },
+    {
+      "epoch": 247.69,
+      "grad_norm": 16.22276496887207,
+      "learning_rate": 1.7437437437437437e-06,
+      "loss": 0.3567,
+      "step": 82480
+    },
+    {
+      "epoch": 247.72,
+      "grad_norm": 12.44990348815918,
+      "learning_rate": 1.742742742742743e-06,
+      "loss": 0.3222,
+      "step": 82490
+    },
+    {
+      "epoch": 247.75,
+      "grad_norm": 10.506450653076172,
+      "learning_rate": 1.7417417417417418e-06,
+      "loss": 0.31,
+      "step": 82500
+    },
+    {
+      "epoch": 247.78,
+      "grad_norm": 16.191402435302734,
+      "learning_rate": 1.740740740740741e-06,
+      "loss": 0.3811,
+      "step": 82510
+    },
+    {
+      "epoch": 247.81,
+      "grad_norm": 12.267305374145508,
+      "learning_rate": 1.7397397397397398e-06,
+      "loss": 0.2822,
+      "step": 82520
+    },
+    {
+      "epoch": 247.84,
+      "grad_norm": 14.286833763122559,
+      "learning_rate": 1.7387387387387388e-06,
+      "loss": 0.3695,
+      "step": 82530
+    },
+    {
+      "epoch": 247.87,
+      "grad_norm": 10.967401504516602,
+      "learning_rate": 1.737737737737738e-06,
+      "loss": 0.3007,
+      "step": 82540
+    },
+    {
+      "epoch": 247.9,
+      "grad_norm": 18.9910945892334,
+      "learning_rate": 1.7367367367367369e-06,
+      "loss": 0.3189,
+      "step": 82550
+    },
+    {
+      "epoch": 247.93,
+      "grad_norm": 11.22144889831543,
+      "learning_rate": 1.735735735735736e-06,
+      "loss": 0.2952,
+      "step": 82560
+    },
+    {
+      "epoch": 247.96,
+      "grad_norm": 15.966447830200195,
+      "learning_rate": 1.734734734734735e-06,
+      "loss": 0.2909,
+      "step": 82570
+    },
+    {
+      "epoch": 247.99,
+      "grad_norm": 21.89383888244629,
+      "learning_rate": 1.733733733733734e-06,
+      "loss": 0.3494,
+      "step": 82580
+    },
+    {
+      "epoch": 248.0,
+      "eval_accuracy": 0.8661,
+      "eval_loss": 0.5498142838478088,
+      "eval_runtime": 12.8086,
+      "eval_samples_per_second": 780.727,
+      "eval_steps_per_second": 3.123,
+      "step": 82584
+    },
+    {
+      "epoch": 248.02,
+      "grad_norm": 15.195852279663086,
+      "learning_rate": 1.732732732732733e-06,
+      "loss": 0.3389,
+      "step": 82590
+    },
+    {
+      "epoch": 248.05,
+      "grad_norm": 14.051218032836914,
+      "learning_rate": 1.731731731731732e-06,
+      "loss": 0.3871,
+      "step": 82600
+    },
+    {
+      "epoch": 248.08,
+      "grad_norm": 11.692228317260742,
+      "learning_rate": 1.7307307307307308e-06,
+      "loss": 0.3397,
+      "step": 82610
+    },
+    {
+      "epoch": 248.11,
+      "grad_norm": 8.283161163330078,
+      "learning_rate": 1.72972972972973e-06,
+      "loss": 0.2958,
+      "step": 82620
+    },
+    {
+      "epoch": 248.14,
+      "grad_norm": 12.822373390197754,
+      "learning_rate": 1.7287287287287288e-06,
+      "loss": 0.3339,
+      "step": 82630
+    },
+    {
+      "epoch": 248.17,
+      "grad_norm": 15.9801664352417,
+      "learning_rate": 1.727727727727728e-06,
+      "loss": 0.3547,
+      "step": 82640
+    },
+    {
+      "epoch": 248.2,
+      "grad_norm": 17.159074783325195,
+      "learning_rate": 1.7267267267267268e-06,
+      "loss": 0.3573,
+      "step": 82650
+    },
+    {
+      "epoch": 248.23,
+      "grad_norm": 26.399452209472656,
+      "learning_rate": 1.7257257257257259e-06,
+      "loss": 0.3496,
+      "step": 82660
+    },
+    {
+      "epoch": 248.26,
+      "grad_norm": 14.603511810302734,
+      "learning_rate": 1.7247247247247249e-06,
+      "loss": 0.33,
+      "step": 82670
+    },
+    {
+      "epoch": 248.29,
+      "grad_norm": 16.983535766601562,
+      "learning_rate": 1.7237237237237239e-06,
+      "loss": 0.3338,
+      "step": 82680
+    },
+    {
+      "epoch": 248.32,
+      "grad_norm": 13.89467716217041,
+      "learning_rate": 1.7227227227227227e-06,
+      "loss": 0.3331,
+      "step": 82690
+    },
+    {
+      "epoch": 248.35,
+      "grad_norm": 12.61744499206543,
+      "learning_rate": 1.721721721721722e-06,
+      "loss": 0.2822,
+      "step": 82700
+    },
+    {
+      "epoch": 248.38,
+      "grad_norm": 11.229924201965332,
+      "learning_rate": 1.7207207207207207e-06,
+      "loss": 0.3626,
+      "step": 82710
+    },
+    {
+      "epoch": 248.41,
+      "grad_norm": 13.064764976501465,
+      "learning_rate": 1.71971971971972e-06,
+      "loss": 0.3311,
+      "step": 82720
+    },
+    {
+      "epoch": 248.44,
+      "grad_norm": 14.491392135620117,
+      "learning_rate": 1.7187187187187188e-06,
+      "loss": 0.2962,
+      "step": 82730
+    },
+    {
+      "epoch": 248.47,
+      "grad_norm": 10.291877746582031,
+      "learning_rate": 1.717717717717718e-06,
+      "loss": 0.3144,
+      "step": 82740
+    },
+    {
+      "epoch": 248.5,
+      "grad_norm": 16.043371200561523,
+      "learning_rate": 1.7167167167167168e-06,
+      "loss": 0.3645,
+      "step": 82750
+    },
+    {
+      "epoch": 248.53,
+      "grad_norm": 16.63316535949707,
+      "learning_rate": 1.7157157157157158e-06,
+      "loss": 0.342,
+      "step": 82760
+    },
+    {
+      "epoch": 248.56,
+      "grad_norm": 15.649832725524902,
+      "learning_rate": 1.7147147147147146e-06,
+      "loss": 0.2803,
+      "step": 82770
+    },
+    {
+      "epoch": 248.59,
+      "grad_norm": 12.151850700378418,
+      "learning_rate": 1.7137137137137139e-06,
+      "loss": 0.327,
+      "step": 82780
+    },
+    {
+      "epoch": 248.62,
+      "grad_norm": 14.1473388671875,
+      "learning_rate": 1.7127127127127127e-06,
+      "loss": 0.3435,
+      "step": 82790
+    },
+    {
+      "epoch": 248.65,
+      "grad_norm": 15.186699867248535,
+      "learning_rate": 1.711711711711712e-06,
+      "loss": 0.3362,
+      "step": 82800
+    },
+    {
+      "epoch": 248.68,
+      "grad_norm": 18.668678283691406,
+      "learning_rate": 1.710710710710711e-06,
+      "loss": 0.3417,
+      "step": 82810
+    },
+    {
+      "epoch": 248.71,
+      "grad_norm": 15.869721412658691,
+      "learning_rate": 1.70970970970971e-06,
+      "loss": 0.2937,
+      "step": 82820
+    },
+    {
+      "epoch": 248.74,
+      "grad_norm": 16.378738403320312,
+      "learning_rate": 1.708708708708709e-06,
+      "loss": 0.386,
+      "step": 82830
+    },
+    {
+      "epoch": 248.77,
+      "grad_norm": 12.948824882507324,
+      "learning_rate": 1.7077077077077078e-06,
+      "loss": 0.3597,
+      "step": 82840
+    },
+    {
+      "epoch": 248.8,
+      "grad_norm": 18.594877243041992,
+      "learning_rate": 1.706706706706707e-06,
+      "loss": 0.3266,
+      "step": 82850
+    },
+    {
+      "epoch": 248.83,
+      "grad_norm": 12.691664695739746,
+      "learning_rate": 1.7057057057057058e-06,
+      "loss": 0.3266,
+      "step": 82860
+    },
+    {
+      "epoch": 248.86,
+      "grad_norm": 30.649314880371094,
+      "learning_rate": 1.704704704704705e-06,
+      "loss": 0.3624,
+      "step": 82870
+    },
+    {
+      "epoch": 248.89,
+      "grad_norm": 16.771272659301758,
+      "learning_rate": 1.7037037037037038e-06,
+      "loss": 0.3697,
+      "step": 82880
+    },
+    {
+      "epoch": 248.92,
+      "grad_norm": 24.083559036254883,
+      "learning_rate": 1.7027027027027028e-06,
+      "loss": 0.3917,
+      "step": 82890
+    },
+    {
+      "epoch": 248.95,
+      "grad_norm": 23.109527587890625,
+      "learning_rate": 1.7017017017017019e-06,
+      "loss": 0.3519,
+      "step": 82900
+    },
+    {
+      "epoch": 248.98,
+      "grad_norm": 14.050040245056152,
+      "learning_rate": 1.7007007007007009e-06,
+      "loss": 0.3543,
+      "step": 82910
+    },
+    {
+      "epoch": 249.0,
+      "eval_accuracy": 0.8652,
+      "eval_loss": 0.5512902140617371,
+      "eval_runtime": 12.6875,
+      "eval_samples_per_second": 788.18,
+      "eval_steps_per_second": 3.153,
+      "step": 82917
+    },
+    {
+      "epoch": 249.01,
+      "grad_norm": 22.97759246826172,
+      "learning_rate": 1.6996996996996997e-06,
+      "loss": 0.3065,
+      "step": 82920
+    },
+    {
+      "epoch": 249.04,
+      "grad_norm": 12.61036205291748,
+      "learning_rate": 1.698698698698699e-06,
+      "loss": 0.3672,
+      "step": 82930
+    },
+    {
+      "epoch": 249.07,
+      "grad_norm": 11.700986862182617,
+      "learning_rate": 1.6976976976976977e-06,
+      "loss": 0.2806,
+      "step": 82940
+    },
+    {
+      "epoch": 249.1,
+      "grad_norm": 12.789227485656738,
+      "learning_rate": 1.696696696696697e-06,
+      "loss": 0.3102,
+      "step": 82950
+    },
+    {
+      "epoch": 249.13,
+      "grad_norm": 20.271902084350586,
+      "learning_rate": 1.6956956956956958e-06,
+      "loss": 0.3522,
+      "step": 82960
+    },
+    {
+      "epoch": 249.16,
+      "grad_norm": 24.179462432861328,
+      "learning_rate": 1.6946946946946948e-06,
+      "loss": 0.2978,
+      "step": 82970
+    },
+    {
+      "epoch": 249.19,
+      "grad_norm": 17.718364715576172,
+      "learning_rate": 1.6936936936936938e-06,
+      "loss": 0.3263,
+      "step": 82980
+    },
+    {
+      "epoch": 249.22,
+      "grad_norm": 16.579402923583984,
+      "learning_rate": 1.6926926926926928e-06,
+      "loss": 0.3082,
+      "step": 82990
+    },
+    {
+      "epoch": 249.25,
+      "grad_norm": 12.037629127502441,
+      "learning_rate": 1.6916916916916916e-06,
+      "loss": 0.3439,
+      "step": 83000
+    },
+    {
+      "epoch": 249.28,
+      "grad_norm": 22.443401336669922,
+      "learning_rate": 1.6906906906906909e-06,
+      "loss": 0.313,
+      "step": 83010
+    },
+    {
+      "epoch": 249.31,
+      "grad_norm": 12.977458953857422,
+      "learning_rate": 1.6896896896896897e-06,
+      "loss": 0.2798,
+      "step": 83020
+    },
+    {
+      "epoch": 249.34,
+      "grad_norm": 24.26066017150879,
+      "learning_rate": 1.6886886886886889e-06,
+      "loss": 0.3831,
+      "step": 83030
+    },
+    {
+      "epoch": 249.37,
+      "grad_norm": 21.303659439086914,
+      "learning_rate": 1.6876876876876877e-06,
+      "loss": 0.3535,
+      "step": 83040
+    },
+    {
+      "epoch": 249.4,
+      "grad_norm": 14.300451278686523,
+      "learning_rate": 1.686686686686687e-06,
+      "loss": 0.3707,
+      "step": 83050
+    },
+    {
+      "epoch": 249.43,
+      "grad_norm": 16.03195571899414,
+      "learning_rate": 1.6856856856856857e-06,
+      "loss": 0.3736,
+      "step": 83060
+    },
+    {
+      "epoch": 249.46,
+      "grad_norm": 16.03146743774414,
+      "learning_rate": 1.6846846846846848e-06,
+      "loss": 0.3017,
+      "step": 83070
+    },
+    {
+      "epoch": 249.49,
+      "grad_norm": 12.4699068069458,
+      "learning_rate": 1.683683683683684e-06,
+      "loss": 0.332,
+      "step": 83080
+    },
+    {
+      "epoch": 249.52,
+      "grad_norm": 13.745859146118164,
+      "learning_rate": 1.6826826826826828e-06,
+      "loss": 0.349,
+      "step": 83090
+    },
+    {
+      "epoch": 249.55,
+      "grad_norm": 10.813240051269531,
+      "learning_rate": 1.681681681681682e-06,
+      "loss": 0.311,
+      "step": 83100
+    },
+    {
+      "epoch": 249.58,
+      "grad_norm": 13.116069793701172,
+      "learning_rate": 1.6806806806806808e-06,
+      "loss": 0.3308,
+      "step": 83110
+    },
+    {
+      "epoch": 249.61,
+      "grad_norm": 12.620058059692383,
+      "learning_rate": 1.6796796796796798e-06,
+      "loss": 0.3411,
+      "step": 83120
+    },
+    {
+      "epoch": 249.64,
+      "grad_norm": 27.88115882873535,
+      "learning_rate": 1.6786786786786789e-06,
+      "loss": 0.3345,
+      "step": 83130
+    },
+    {
+      "epoch": 249.67,
+      "grad_norm": 16.893930435180664,
+      "learning_rate": 1.6776776776776779e-06,
+      "loss": 0.3582,
+      "step": 83140
+    },
+    {
+      "epoch": 249.7,
+      "grad_norm": 23.508281707763672,
+      "learning_rate": 1.6766766766766767e-06,
+      "loss": 0.3612,
+      "step": 83150
+    },
+    {
+      "epoch": 249.73,
+      "grad_norm": 12.831755638122559,
+      "learning_rate": 1.675675675675676e-06,
+      "loss": 0.3711,
+      "step": 83160
+    },
+    {
+      "epoch": 249.76,
+      "grad_norm": 10.547800064086914,
+      "learning_rate": 1.6746746746746747e-06,
+      "loss": 0.3524,
+      "step": 83170
+    },
+    {
+      "epoch": 249.79,
+      "grad_norm": 15.264772415161133,
+      "learning_rate": 1.673673673673674e-06,
+      "loss": 0.3165,
+      "step": 83180
+    },
+    {
+      "epoch": 249.82,
+      "grad_norm": 16.811094284057617,
+      "learning_rate": 1.6726726726726728e-06,
+      "loss": 0.309,
+      "step": 83190
+    },
+    {
+      "epoch": 249.85,
+      "grad_norm": 15.843195915222168,
+      "learning_rate": 1.6716716716716718e-06,
+      "loss": 0.354,
+      "step": 83200
+    },
+    {
+      "epoch": 249.88,
+      "grad_norm": 21.120059967041016,
+      "learning_rate": 1.6706706706706708e-06,
+      "loss": 0.3298,
+      "step": 83210
+    },
+    {
+      "epoch": 249.91,
+      "grad_norm": 13.018309593200684,
+      "learning_rate": 1.6696696696696698e-06,
+      "loss": 0.2983,
+      "step": 83220
+    },
+    {
+      "epoch": 249.94,
+      "grad_norm": 25.62592124938965,
+      "learning_rate": 1.6686686686686686e-06,
+      "loss": 0.3599,
+      "step": 83230
+    },
+    {
+      "epoch": 249.97,
+      "grad_norm": 13.399046897888184,
+      "learning_rate": 1.6676676676676679e-06,
+      "loss": 0.3416,
+      "step": 83240
+    },
+    {
+      "epoch": 250.0,
+      "grad_norm": 0.35038629174232483,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.3142,
+      "step": 83250
+    },
+    {
+      "epoch": 250.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5506334900856018,
+      "eval_runtime": 12.6163,
+      "eval_samples_per_second": 792.628,
+      "eval_steps_per_second": 3.171,
+      "step": 83250
+    },
+    {
+      "epoch": 250.03,
+      "grad_norm": 18.12055206298828,
+      "learning_rate": 1.6656656656656659e-06,
+      "loss": 0.3283,
+      "step": 83260
+    },
+    {
+      "epoch": 250.06,
+      "grad_norm": 14.023235321044922,
+      "learning_rate": 1.6646646646646647e-06,
+      "loss": 0.3145,
+      "step": 83270
+    },
+    {
+      "epoch": 250.09,
+      "grad_norm": 15.472148895263672,
+      "learning_rate": 1.663663663663664e-06,
+      "loss": 0.2968,
+      "step": 83280
+    },
+    {
+      "epoch": 250.12,
+      "grad_norm": 15.165860176086426,
+      "learning_rate": 1.6626626626626627e-06,
+      "loss": 0.3486,
+      "step": 83290
+    },
+    {
+      "epoch": 250.15,
+      "grad_norm": 20.496858596801758,
+      "learning_rate": 1.6616616616616618e-06,
+      "loss": 0.3115,
+      "step": 83300
+    },
+    {
+      "epoch": 250.18,
+      "grad_norm": 19.575021743774414,
+      "learning_rate": 1.6606606606606608e-06,
+      "loss": 0.3326,
+      "step": 83310
+    },
+    {
+      "epoch": 250.21,
+      "grad_norm": 16.95491600036621,
+      "learning_rate": 1.6596596596596598e-06,
+      "loss": 0.3656,
+      "step": 83320
+    },
+    {
+      "epoch": 250.24,
+      "grad_norm": 17.2111873626709,
+      "learning_rate": 1.6586586586586586e-06,
+      "loss": 0.3671,
+      "step": 83330
+    },
+    {
+      "epoch": 250.27,
+      "grad_norm": 24.39700698852539,
+      "learning_rate": 1.6576576576576578e-06,
+      "loss": 0.3142,
+      "step": 83340
+    },
+    {
+      "epoch": 250.3,
+      "grad_norm": 11.729545593261719,
+      "learning_rate": 1.6566566566566568e-06,
+      "loss": 0.3504,
+      "step": 83350
+    },
+    {
+      "epoch": 250.33,
+      "grad_norm": 14.64574909210205,
+      "learning_rate": 1.6556556556556559e-06,
+      "loss": 0.3108,
+      "step": 83360
+    },
+    {
+      "epoch": 250.36,
+      "grad_norm": 15.491724967956543,
+      "learning_rate": 1.6546546546546549e-06,
+      "loss": 0.3121,
+      "step": 83370
+    },
+    {
+      "epoch": 250.39,
+      "grad_norm": 20.375911712646484,
+      "learning_rate": 1.6536536536536537e-06,
+      "loss": 0.3558,
+      "step": 83380
+    },
+    {
+      "epoch": 250.42,
+      "grad_norm": 12.061564445495605,
+      "learning_rate": 1.652652652652653e-06,
+      "loss": 0.3273,
+      "step": 83390
+    },
+    {
+      "epoch": 250.45,
+      "grad_norm": 13.83393669128418,
+      "learning_rate": 1.6516516516516517e-06,
+      "loss": 0.389,
+      "step": 83400
+    },
+    {
+      "epoch": 250.48,
+      "grad_norm": 10.640325546264648,
+      "learning_rate": 1.650650650650651e-06,
+      "loss": 0.326,
+      "step": 83410
+    },
+    {
+      "epoch": 250.51,
+      "grad_norm": 13.880182266235352,
+      "learning_rate": 1.6496496496496498e-06,
+      "loss": 0.3376,
+      "step": 83420
+    },
+    {
+      "epoch": 250.54,
+      "grad_norm": 13.632920265197754,
+      "learning_rate": 1.6486486486486488e-06,
+      "loss": 0.3498,
+      "step": 83430
+    },
+    {
+      "epoch": 250.57,
+      "grad_norm": 10.799054145812988,
+      "learning_rate": 1.6476476476476478e-06,
+      "loss": 0.3346,
+      "step": 83440
+    },
+    {
+      "epoch": 250.6,
+      "grad_norm": 13.45500373840332,
+      "learning_rate": 1.6466466466466468e-06,
+      "loss": 0.3461,
+      "step": 83450
+    },
+    {
+      "epoch": 250.63,
+      "grad_norm": 18.058856964111328,
+      "learning_rate": 1.6456456456456456e-06,
+      "loss": 0.3265,
+      "step": 83460
+    },
+    {
+      "epoch": 250.66,
+      "grad_norm": 20.532861709594727,
+      "learning_rate": 1.6446446446446449e-06,
+      "loss": 0.3209,
+      "step": 83470
+    },
+    {
+      "epoch": 250.69,
+      "grad_norm": 19.048017501831055,
+      "learning_rate": 1.6436436436436437e-06,
+      "loss": 0.3532,
+      "step": 83480
+    },
+    {
+      "epoch": 250.72,
+      "grad_norm": 20.907899856567383,
+      "learning_rate": 1.6426426426426429e-06,
+      "loss": 0.3499,
+      "step": 83490
+    },
+    {
+      "epoch": 250.75,
+      "grad_norm": 13.779293060302734,
+      "learning_rate": 1.6416416416416417e-06,
+      "loss": 0.3297,
+      "step": 83500
+    },
+    {
+      "epoch": 250.78,
+      "grad_norm": 12.73255729675293,
+      "learning_rate": 1.640640640640641e-06,
+      "loss": 0.2961,
+      "step": 83510
+    },
+    {
+      "epoch": 250.81,
+      "grad_norm": 12.403632164001465,
+      "learning_rate": 1.6396396396396397e-06,
+      "loss": 0.298,
+      "step": 83520
+    },
+    {
+      "epoch": 250.84,
+      "grad_norm": 15.160212516784668,
+      "learning_rate": 1.6386386386386388e-06,
+      "loss": 0.3505,
+      "step": 83530
+    },
+    {
+      "epoch": 250.87,
+      "grad_norm": 14.740659713745117,
+      "learning_rate": 1.6376376376376378e-06,
+      "loss": 0.3336,
+      "step": 83540
+    },
+    {
+      "epoch": 250.9,
+      "grad_norm": 14.701874732971191,
+      "learning_rate": 1.6366366366366368e-06,
+      "loss": 0.3293,
+      "step": 83550
+    },
+    {
+      "epoch": 250.93,
+      "grad_norm": 20.26247215270996,
+      "learning_rate": 1.6356356356356356e-06,
+      "loss": 0.3814,
+      "step": 83560
+    },
+    {
+      "epoch": 250.96,
+      "grad_norm": 17.141246795654297,
+      "learning_rate": 1.6346346346346348e-06,
+      "loss": 0.352,
+      "step": 83570
+    },
+    {
+      "epoch": 250.99,
+      "grad_norm": 12.621345520019531,
+      "learning_rate": 1.6336336336336336e-06,
+      "loss": 0.3534,
+      "step": 83580
+    },
+    {
+      "epoch": 251.0,
+      "eval_accuracy": 0.8657,
+      "eval_loss": 0.5471351146697998,
+      "eval_runtime": 12.7269,
+      "eval_samples_per_second": 785.74,
+      "eval_steps_per_second": 3.143,
+      "step": 83583
+    },
+    {
+      "epoch": 251.02,
+      "grad_norm": 17.02467918395996,
+      "learning_rate": 1.6326326326326329e-06,
+      "loss": 0.3211,
+      "step": 83590
+    },
+    {
+      "epoch": 251.05,
+      "grad_norm": 13.94760513305664,
+      "learning_rate": 1.6316316316316317e-06,
+      "loss": 0.2896,
+      "step": 83600
+    },
+    {
+      "epoch": 251.08,
+      "grad_norm": 30.221357345581055,
+      "learning_rate": 1.6306306306306307e-06,
+      "loss": 0.3445,
+      "step": 83610
+    },
+    {
+      "epoch": 251.11,
+      "grad_norm": 15.346515655517578,
+      "learning_rate": 1.62962962962963e-06,
+      "loss": 0.3553,
+      "step": 83620
+    },
+    {
+      "epoch": 251.14,
+      "grad_norm": 16.47450828552246,
+      "learning_rate": 1.6286286286286287e-06,
+      "loss": 0.3049,
+      "step": 83630
+    },
+    {
+      "epoch": 251.17,
+      "grad_norm": 21.740264892578125,
+      "learning_rate": 1.627627627627628e-06,
+      "loss": 0.3261,
+      "step": 83640
+    },
+    {
+      "epoch": 251.2,
+      "grad_norm": 15.965238571166992,
+      "learning_rate": 1.6266266266266268e-06,
+      "loss": 0.2999,
+      "step": 83650
+    },
+    {
+      "epoch": 251.23,
+      "grad_norm": 17.615345001220703,
+      "learning_rate": 1.6256256256256258e-06,
+      "loss": 0.3228,
+      "step": 83660
+    },
+    {
+      "epoch": 251.26,
+      "grad_norm": 16.830402374267578,
+      "learning_rate": 1.6246246246246248e-06,
+      "loss": 0.3329,
+      "step": 83670
+    },
+    {
+      "epoch": 251.29,
+      "grad_norm": 16.412303924560547,
+      "learning_rate": 1.6236236236236238e-06,
+      "loss": 0.3237,
+      "step": 83680
+    },
+    {
+      "epoch": 251.32,
+      "grad_norm": 11.32098388671875,
+      "learning_rate": 1.6226226226226226e-06,
+      "loss": 0.272,
+      "step": 83690
+    },
+    {
+      "epoch": 251.35,
+      "grad_norm": 14.871081352233887,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.2864,
+      "step": 83700
+    },
+    {
+      "epoch": 251.38,
+      "grad_norm": 20.095212936401367,
+      "learning_rate": 1.6206206206206207e-06,
+      "loss": 0.3462,
+      "step": 83710
+    },
+    {
+      "epoch": 251.41,
+      "grad_norm": 18.96649742126465,
+      "learning_rate": 1.6196196196196199e-06,
+      "loss": 0.3368,
+      "step": 83720
+    },
+    {
+      "epoch": 251.44,
+      "grad_norm": 9.351943016052246,
+      "learning_rate": 1.6186186186186187e-06,
+      "loss": 0.3228,
+      "step": 83730
+    },
+    {
+      "epoch": 251.47,
+      "grad_norm": 14.232733726501465,
+      "learning_rate": 1.617617617617618e-06,
+      "loss": 0.2913,
+      "step": 83740
+    },
+    {
+      "epoch": 251.5,
+      "grad_norm": 11.356451988220215,
+      "learning_rate": 1.6166166166166167e-06,
+      "loss": 0.3009,
+      "step": 83750
+    },
+    {
+      "epoch": 251.53,
+      "grad_norm": 10.036985397338867,
+      "learning_rate": 1.6156156156156157e-06,
+      "loss": 0.3284,
+      "step": 83760
+    },
+    {
+      "epoch": 251.56,
+      "grad_norm": 19.136987686157227,
+      "learning_rate": 1.6146146146146146e-06,
+      "loss": 0.312,
+      "step": 83770
+    },
+    {
+      "epoch": 251.59,
+      "grad_norm": 12.332791328430176,
+      "learning_rate": 1.6136136136136138e-06,
+      "loss": 0.3505,
+      "step": 83780
+    },
+    {
+      "epoch": 251.62,
+      "grad_norm": 12.47145938873291,
+      "learning_rate": 1.6126126126126126e-06,
+      "loss": 0.3334,
+      "step": 83790
+    },
+    {
+      "epoch": 251.65,
+      "grad_norm": 18.086713790893555,
+      "learning_rate": 1.6116116116116118e-06,
+      "loss": 0.3376,
+      "step": 83800
+    },
+    {
+      "epoch": 251.68,
+      "grad_norm": 20.126102447509766,
+      "learning_rate": 1.6106106106106106e-06,
+      "loss": 0.3618,
+      "step": 83810
+    },
+    {
+      "epoch": 251.71,
+      "grad_norm": 19.111431121826172,
+      "learning_rate": 1.6096096096096099e-06,
+      "loss": 0.3471,
+      "step": 83820
+    },
+    {
+      "epoch": 251.74,
+      "grad_norm": 9.603525161743164,
+      "learning_rate": 1.6086086086086087e-06,
+      "loss": 0.2864,
+      "step": 83830
+    },
+    {
+      "epoch": 251.77,
+      "grad_norm": 14.761211395263672,
+      "learning_rate": 1.6076076076076077e-06,
+      "loss": 0.3344,
+      "step": 83840
+    },
+    {
+      "epoch": 251.8,
+      "grad_norm": 20.331083297729492,
+      "learning_rate": 1.6066066066066067e-06,
+      "loss": 0.3443,
+      "step": 83850
+    },
+    {
+      "epoch": 251.83,
+      "grad_norm": 12.775331497192383,
+      "learning_rate": 1.6056056056056057e-06,
+      "loss": 0.3391,
+      "step": 83860
+    },
+    {
+      "epoch": 251.86,
+      "grad_norm": 18.374711990356445,
+      "learning_rate": 1.6046046046046045e-06,
+      "loss": 0.3358,
+      "step": 83870
+    },
+    {
+      "epoch": 251.89,
+      "grad_norm": 19.453798294067383,
+      "learning_rate": 1.6036036036036038e-06,
+      "loss": 0.3771,
+      "step": 83880
+    },
+    {
+      "epoch": 251.92,
+      "grad_norm": 16.072202682495117,
+      "learning_rate": 1.6026026026026028e-06,
+      "loss": 0.3654,
+      "step": 83890
+    },
+    {
+      "epoch": 251.95,
+      "grad_norm": 14.802956581115723,
+      "learning_rate": 1.6016016016016018e-06,
+      "loss": 0.3365,
+      "step": 83900
+    },
+    {
+      "epoch": 251.98,
+      "grad_norm": 17.33298110961914,
+      "learning_rate": 1.6006006006006008e-06,
+      "loss": 0.3703,
+      "step": 83910
+    },
+    {
+      "epoch": 252.0,
+      "eval_accuracy": 0.866,
+      "eval_loss": 0.5474722385406494,
+      "eval_runtime": 12.8096,
+      "eval_samples_per_second": 780.662,
+      "eval_steps_per_second": 3.123,
+      "step": 83916
+    },
+    {
+      "epoch": 252.01,
+      "grad_norm": 12.429125785827637,
+      "learning_rate": 1.5995995995995996e-06,
+      "loss": 0.2794,
+      "step": 83920
+    },
+    {
+      "epoch": 252.04,
+      "grad_norm": 17.728761672973633,
+      "learning_rate": 1.5985985985985988e-06,
+      "loss": 0.3279,
+      "step": 83930
+    },
+    {
+      "epoch": 252.07,
+      "grad_norm": 20.128942489624023,
+      "learning_rate": 1.5975975975975977e-06,
+      "loss": 0.4021,
+      "step": 83940
+    },
+    {
+      "epoch": 252.1,
+      "grad_norm": 18.31165313720703,
+      "learning_rate": 1.5965965965965969e-06,
+      "loss": 0.3182,
+      "step": 83950
+    },
+    {
+      "epoch": 252.13,
+      "grad_norm": 9.864901542663574,
+      "learning_rate": 1.5955955955955957e-06,
+      "loss": 0.2977,
+      "step": 83960
+    },
+    {
+      "epoch": 252.16,
+      "grad_norm": 10.610376358032227,
+      "learning_rate": 1.5945945945945947e-06,
+      "loss": 0.3395,
+      "step": 83970
+    },
+    {
+      "epoch": 252.19,
+      "grad_norm": 19.28207015991211,
+      "learning_rate": 1.5935935935935937e-06,
+      "loss": 0.2946,
+      "step": 83980
+    },
+    {
+      "epoch": 252.22,
+      "grad_norm": 18.509552001953125,
+      "learning_rate": 1.5925925925925927e-06,
+      "loss": 0.3951,
+      "step": 83990
+    },
+    {
+      "epoch": 252.25,
+      "grad_norm": 16.97201919555664,
+      "learning_rate": 1.5915915915915916e-06,
+      "loss": 0.2875,
+      "step": 84000
+    },
+    {
+      "epoch": 252.28,
+      "grad_norm": 23.649099349975586,
+      "learning_rate": 1.5905905905905908e-06,
+      "loss": 0.3496,
+      "step": 84010
+    },
+    {
+      "epoch": 252.31,
+      "grad_norm": 23.15732765197754,
+      "learning_rate": 1.5895895895895896e-06,
+      "loss": 0.3332,
+      "step": 84020
+    },
+    {
+      "epoch": 252.34,
+      "grad_norm": 9.025938034057617,
+      "learning_rate": 1.5885885885885888e-06,
+      "loss": 0.3418,
+      "step": 84030
+    },
+    {
+      "epoch": 252.37,
+      "grad_norm": 14.576842308044434,
+      "learning_rate": 1.5875875875875876e-06,
+      "loss": 0.3479,
+      "step": 84040
+    },
+    {
+      "epoch": 252.4,
+      "grad_norm": 9.807663917541504,
+      "learning_rate": 1.5865865865865869e-06,
+      "loss": 0.356,
+      "step": 84050
+    },
+    {
+      "epoch": 252.43,
+      "grad_norm": 9.558355331420898,
+      "learning_rate": 1.5855855855855857e-06,
+      "loss": 0.3392,
+      "step": 84060
+    },
+    {
+      "epoch": 252.46,
+      "grad_norm": 8.781988143920898,
+      "learning_rate": 1.5845845845845847e-06,
+      "loss": 0.3074,
+      "step": 84070
+    },
+    {
+      "epoch": 252.49,
+      "grad_norm": 13.420248985290527,
+      "learning_rate": 1.5835835835835837e-06,
+      "loss": 0.3309,
+      "step": 84080
+    },
+    {
+      "epoch": 252.52,
+      "grad_norm": 21.785049438476562,
+      "learning_rate": 1.5825825825825827e-06,
+      "loss": 0.3003,
+      "step": 84090
+    },
+    {
+      "epoch": 252.55,
+      "grad_norm": 11.656575202941895,
+      "learning_rate": 1.5815815815815815e-06,
+      "loss": 0.3534,
+      "step": 84100
+    },
+    {
+      "epoch": 252.58,
+      "grad_norm": 16.7396183013916,
+      "learning_rate": 1.5805805805805808e-06,
+      "loss": 0.361,
+      "step": 84110
+    },
+    {
+      "epoch": 252.61,
+      "grad_norm": 14.75139331817627,
+      "learning_rate": 1.5795795795795796e-06,
+      "loss": 0.3239,
+      "step": 84120
+    },
+    {
+      "epoch": 252.64,
+      "grad_norm": 21.33076286315918,
+      "learning_rate": 1.5785785785785788e-06,
+      "loss": 0.3269,
+      "step": 84130
+    },
+    {
+      "epoch": 252.67,
+      "grad_norm": 12.162034034729004,
+      "learning_rate": 1.5775775775775776e-06,
+      "loss": 0.3049,
+      "step": 84140
+    },
+    {
+      "epoch": 252.7,
+      "grad_norm": 16.561504364013672,
+      "learning_rate": 1.5765765765765766e-06,
+      "loss": 0.3644,
+      "step": 84150
+    },
+    {
+      "epoch": 252.73,
+      "grad_norm": 18.392330169677734,
+      "learning_rate": 1.5755755755755758e-06,
+      "loss": 0.292,
+      "step": 84160
+    },
+    {
+      "epoch": 252.76,
+      "grad_norm": 8.75645923614502,
+      "learning_rate": 1.5745745745745747e-06,
+      "loss": 0.3341,
+      "step": 84170
+    },
+    {
+      "epoch": 252.79,
+      "grad_norm": 10.35513973236084,
+      "learning_rate": 1.5735735735735739e-06,
+      "loss": 0.3184,
+      "step": 84180
+    },
+    {
+      "epoch": 252.82,
+      "grad_norm": 14.366315841674805,
+      "learning_rate": 1.5725725725725727e-06,
+      "loss": 0.2981,
+      "step": 84190
+    },
+    {
+      "epoch": 252.85,
+      "grad_norm": 48.18925476074219,
+      "learning_rate": 1.5715715715715717e-06,
+      "loss": 0.3204,
+      "step": 84200
+    },
+    {
+      "epoch": 252.88,
+      "grad_norm": 21.52358055114746,
+      "learning_rate": 1.5705705705705707e-06,
+      "loss": 0.3047,
+      "step": 84210
+    },
+    {
+      "epoch": 252.91,
+      "grad_norm": 27.66925048828125,
+      "learning_rate": 1.5695695695695697e-06,
+      "loss": 0.2912,
+      "step": 84220
+    },
+    {
+      "epoch": 252.94,
+      "grad_norm": 14.87544059753418,
+      "learning_rate": 1.5685685685685686e-06,
+      "loss": 0.3085,
+      "step": 84230
+    },
+    {
+      "epoch": 252.97,
+      "grad_norm": 19.86974334716797,
+      "learning_rate": 1.5675675675675678e-06,
+      "loss": 0.3331,
+      "step": 84240
+    },
+    {
+      "epoch": 253.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.5462300777435303,
+      "eval_runtime": 12.8022,
+      "eval_samples_per_second": 781.118,
+      "eval_steps_per_second": 3.124,
+      "step": 84249
+    },
+    {
+      "epoch": 253.0,
+      "grad_norm": 13.862272262573242,
+      "learning_rate": 1.5665665665665666e-06,
+      "loss": 0.3079,
+      "step": 84250
+    },
+    {
+      "epoch": 253.03,
+      "grad_norm": 13.843643188476562,
+      "learning_rate": 1.5655655655655658e-06,
+      "loss": 0.3211,
+      "step": 84260
+    },
+    {
+      "epoch": 253.06,
+      "grad_norm": 9.771224975585938,
+      "learning_rate": 1.5645645645645646e-06,
+      "loss": 0.3315,
+      "step": 84270
+    },
+    {
+      "epoch": 253.09,
+      "grad_norm": 13.968844413757324,
+      "learning_rate": 1.5635635635635639e-06,
+      "loss": 0.3315,
+      "step": 84280
+    },
+    {
+      "epoch": 253.12,
+      "grad_norm": 24.386316299438477,
+      "learning_rate": 1.5625625625625627e-06,
+      "loss": 0.3642,
+      "step": 84290
+    },
+    {
+      "epoch": 253.15,
+      "grad_norm": 12.079916954040527,
+      "learning_rate": 1.5615615615615617e-06,
+      "loss": 0.3412,
+      "step": 84300
+    },
+    {
+      "epoch": 253.18,
+      "grad_norm": 20.331998825073242,
+      "learning_rate": 1.5605605605605607e-06,
+      "loss": 0.3087,
+      "step": 84310
+    },
+    {
+      "epoch": 253.21,
+      "grad_norm": 21.449691772460938,
+      "learning_rate": 1.5595595595595597e-06,
+      "loss": 0.3432,
+      "step": 84320
+    },
+    {
+      "epoch": 253.24,
+      "grad_norm": 14.867684364318848,
+      "learning_rate": 1.5585585585585585e-06,
+      "loss": 0.3504,
+      "step": 84330
+    },
+    {
+      "epoch": 253.27,
+      "grad_norm": 17.501007080078125,
+      "learning_rate": 1.5575575575575578e-06,
+      "loss": 0.3543,
+      "step": 84340
+    },
+    {
+      "epoch": 253.3,
+      "grad_norm": 19.008268356323242,
+      "learning_rate": 1.5565565565565566e-06,
+      "loss": 0.3618,
+      "step": 84350
+    },
+    {
+      "epoch": 253.33,
+      "grad_norm": 18.469318389892578,
+      "learning_rate": 1.5555555555555558e-06,
+      "loss": 0.3897,
+      "step": 84360
+    },
+    {
+      "epoch": 253.36,
+      "grad_norm": 12.642078399658203,
+      "learning_rate": 1.5545545545545546e-06,
+      "loss": 0.3148,
+      "step": 84370
+    },
+    {
+      "epoch": 253.39,
+      "grad_norm": 20.082021713256836,
+      "learning_rate": 1.5535535535535536e-06,
+      "loss": 0.3326,
+      "step": 84380
+    },
+    {
+      "epoch": 253.42,
+      "grad_norm": 16.924890518188477,
+      "learning_rate": 1.5525525525525526e-06,
+      "loss": 0.4082,
+      "step": 84390
+    },
+    {
+      "epoch": 253.45,
+      "grad_norm": 25.237773895263672,
+      "learning_rate": 1.5515515515515517e-06,
+      "loss": 0.3483,
+      "step": 84400
+    },
+    {
+      "epoch": 253.48,
+      "grad_norm": 24.62592887878418,
+      "learning_rate": 1.5505505505505509e-06,
+      "loss": 0.3231,
+      "step": 84410
+    },
+    {
+      "epoch": 253.51,
+      "grad_norm": 14.649287223815918,
+      "learning_rate": 1.5495495495495497e-06,
+      "loss": 0.3414,
+      "step": 84420
+    },
+    {
+      "epoch": 253.54,
+      "grad_norm": 16.678482055664062,
+      "learning_rate": 1.5485485485485487e-06,
+      "loss": 0.402,
+      "step": 84430
+    },
+    {
+      "epoch": 253.57,
+      "grad_norm": 16.042438507080078,
+      "learning_rate": 1.5475475475475477e-06,
+      "loss": 0.3861,
+      "step": 84440
+    },
+    {
+      "epoch": 253.6,
+      "grad_norm": 12.944937705993652,
+      "learning_rate": 1.5465465465465467e-06,
+      "loss": 0.3498,
+      "step": 84450
+    },
+    {
+      "epoch": 253.63,
+      "grad_norm": 17.321271896362305,
+      "learning_rate": 1.5455455455455456e-06,
+      "loss": 0.3155,
+      "step": 84460
+    },
+    {
+      "epoch": 253.66,
+      "grad_norm": 22.769649505615234,
+      "learning_rate": 1.5445445445445448e-06,
+      "loss": 0.3705,
+      "step": 84470
+    },
+    {
+      "epoch": 253.69,
+      "grad_norm": 13.093915939331055,
+      "learning_rate": 1.5435435435435436e-06,
+      "loss": 0.3111,
+      "step": 84480
+    },
+    {
+      "epoch": 253.72,
+      "grad_norm": 13.849239349365234,
+      "learning_rate": 1.5425425425425428e-06,
+      "loss": 0.294,
+      "step": 84490
+    },
+    {
+      "epoch": 253.75,
+      "grad_norm": 10.011273384094238,
+      "learning_rate": 1.5415415415415416e-06,
+      "loss": 0.3025,
+      "step": 84500
+    },
+    {
+      "epoch": 253.78,
+      "grad_norm": 16.47110939025879,
+      "learning_rate": 1.5405405405405409e-06,
+      "loss": 0.347,
+      "step": 84510
+    },
+    {
+      "epoch": 253.81,
+      "grad_norm": 18.235862731933594,
+      "learning_rate": 1.5395395395395397e-06,
+      "loss": 0.3011,
+      "step": 84520
+    },
+    {
+      "epoch": 253.84,
+      "grad_norm": 8.767010688781738,
+      "learning_rate": 1.5385385385385387e-06,
+      "loss": 0.3458,
+      "step": 84530
+    },
+    {
+      "epoch": 253.87,
+      "grad_norm": 18.652244567871094,
+      "learning_rate": 1.5375375375375377e-06,
+      "loss": 0.3638,
+      "step": 84540
+    },
+    {
+      "epoch": 253.9,
+      "grad_norm": 15.755702018737793,
+      "learning_rate": 1.5365365365365367e-06,
+      "loss": 0.3675,
+      "step": 84550
+    },
+    {
+      "epoch": 253.93,
+      "grad_norm": 11.600528717041016,
+      "learning_rate": 1.5355355355355355e-06,
+      "loss": 0.2839,
+      "step": 84560
+    },
+    {
+      "epoch": 253.96,
+      "grad_norm": 15.53160572052002,
+      "learning_rate": 1.5345345345345348e-06,
+      "loss": 0.342,
+      "step": 84570
+    },
+    {
+      "epoch": 253.99,
+      "grad_norm": 17.484195709228516,
+      "learning_rate": 1.5335335335335336e-06,
+      "loss": 0.3349,
+      "step": 84580
+    },
+    {
+      "epoch": 254.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5466774106025696,
+      "eval_runtime": 12.827,
+      "eval_samples_per_second": 779.606,
+      "eval_steps_per_second": 3.118,
+      "step": 84582
+    },
+    {
+      "epoch": 254.02,
+      "grad_norm": 18.835363388061523,
+      "learning_rate": 1.5325325325325328e-06,
+      "loss": 0.3252,
+      "step": 84590
+    },
+    {
+      "epoch": 254.05,
+      "grad_norm": 24.752824783325195,
+      "learning_rate": 1.5315315315315316e-06,
+      "loss": 0.328,
+      "step": 84600
+    },
+    {
+      "epoch": 254.08,
+      "grad_norm": 15.009206771850586,
+      "learning_rate": 1.5305305305305306e-06,
+      "loss": 0.3015,
+      "step": 84610
+    },
+    {
+      "epoch": 254.11,
+      "grad_norm": 16.529390335083008,
+      "learning_rate": 1.5295295295295296e-06,
+      "loss": 0.314,
+      "step": 84620
+    },
+    {
+      "epoch": 254.14,
+      "grad_norm": 14.637746810913086,
+      "learning_rate": 1.5285285285285287e-06,
+      "loss": 0.3466,
+      "step": 84630
+    },
+    {
+      "epoch": 254.17,
+      "grad_norm": 23.961750030517578,
+      "learning_rate": 1.5275275275275275e-06,
+      "loss": 0.2654,
+      "step": 84640
+    },
+    {
+      "epoch": 254.2,
+      "grad_norm": 21.756362915039062,
+      "learning_rate": 1.5265265265265267e-06,
+      "loss": 0.3299,
+      "step": 84650
+    },
+    {
+      "epoch": 254.23,
+      "grad_norm": 8.522087097167969,
+      "learning_rate": 1.5255255255255255e-06,
+      "loss": 0.3231,
+      "step": 84660
+    },
+    {
+      "epoch": 254.26,
+      "grad_norm": 26.567827224731445,
+      "learning_rate": 1.5245245245245247e-06,
+      "loss": 0.3544,
+      "step": 84670
+    },
+    {
+      "epoch": 254.29,
+      "grad_norm": 13.041289329528809,
+      "learning_rate": 1.5235235235235237e-06,
+      "loss": 0.2864,
+      "step": 84680
+    },
+    {
+      "epoch": 254.32,
+      "grad_norm": 20.45798110961914,
+      "learning_rate": 1.5225225225225225e-06,
+      "loss": 0.3761,
+      "step": 84690
+    },
+    {
+      "epoch": 254.35,
+      "grad_norm": 19.165618896484375,
+      "learning_rate": 1.5215215215215218e-06,
+      "loss": 0.3231,
+      "step": 84700
+    },
+    {
+      "epoch": 254.38,
+      "grad_norm": 22.511363983154297,
+      "learning_rate": 1.5205205205205206e-06,
+      "loss": 0.3489,
+      "step": 84710
+    },
+    {
+      "epoch": 254.41,
+      "grad_norm": 21.799991607666016,
+      "learning_rate": 1.5195195195195198e-06,
+      "loss": 0.3351,
+      "step": 84720
+    },
+    {
+      "epoch": 254.44,
+      "grad_norm": 12.351995468139648,
+      "learning_rate": 1.5185185185185186e-06,
+      "loss": 0.3123,
+      "step": 84730
+    },
+    {
+      "epoch": 254.47,
+      "grad_norm": 15.831256866455078,
+      "learning_rate": 1.5175175175175176e-06,
+      "loss": 0.2741,
+      "step": 84740
+    },
+    {
+      "epoch": 254.5,
+      "grad_norm": 18.95428466796875,
+      "learning_rate": 1.5165165165165167e-06,
+      "loss": 0.3087,
+      "step": 84750
+    },
+    {
+      "epoch": 254.53,
+      "grad_norm": 12.525662422180176,
+      "learning_rate": 1.5155155155155157e-06,
+      "loss": 0.337,
+      "step": 84760
+    },
+    {
+      "epoch": 254.56,
+      "grad_norm": 26.797147750854492,
+      "learning_rate": 1.5145145145145145e-06,
+      "loss": 0.3013,
+      "step": 84770
+    },
+    {
+      "epoch": 254.59,
+      "grad_norm": 12.378711700439453,
+      "learning_rate": 1.5135135135135137e-06,
+      "loss": 0.3235,
+      "step": 84780
+    },
+    {
+      "epoch": 254.62,
+      "grad_norm": 10.917810440063477,
+      "learning_rate": 1.5125125125125125e-06,
+      "loss": 0.2873,
+      "step": 84790
+    },
+    {
+      "epoch": 254.65,
+      "grad_norm": 16.842288970947266,
+      "learning_rate": 1.5115115115115118e-06,
+      "loss": 0.3051,
+      "step": 84800
+    },
+    {
+      "epoch": 254.68,
+      "grad_norm": 14.727993965148926,
+      "learning_rate": 1.5105105105105106e-06,
+      "loss": 0.3003,
+      "step": 84810
+    },
+    {
+      "epoch": 254.71,
+      "grad_norm": 13.320401191711426,
+      "learning_rate": 1.5095095095095098e-06,
+      "loss": 0.3399,
+      "step": 84820
+    },
+    {
+      "epoch": 254.74,
+      "grad_norm": 13.441773414611816,
+      "learning_rate": 1.5085085085085086e-06,
+      "loss": 0.379,
+      "step": 84830
+    },
+    {
+      "epoch": 254.77,
+      "grad_norm": 19.164329528808594,
+      "learning_rate": 1.5075075075075076e-06,
+      "loss": 0.3321,
+      "step": 84840
+    },
+    {
+      "epoch": 254.8,
+      "grad_norm": 14.655036926269531,
+      "learning_rate": 1.5065065065065066e-06,
+      "loss": 0.3588,
+      "step": 84850
+    },
+    {
+      "epoch": 254.83,
+      "grad_norm": 19.923921585083008,
+      "learning_rate": 1.5055055055055056e-06,
+      "loss": 0.3096,
+      "step": 84860
+    },
+    {
+      "epoch": 254.86,
+      "grad_norm": 15.240574836730957,
+      "learning_rate": 1.5045045045045045e-06,
+      "loss": 0.3726,
+      "step": 84870
+    },
+    {
+      "epoch": 254.89,
+      "grad_norm": 11.486536979675293,
+      "learning_rate": 1.5035035035035037e-06,
+      "loss": 0.307,
+      "step": 84880
+    },
+    {
+      "epoch": 254.92,
+      "grad_norm": 10.147424697875977,
+      "learning_rate": 1.5025025025025025e-06,
+      "loss": 0.3067,
+      "step": 84890
+    },
+    {
+      "epoch": 254.95,
+      "grad_norm": 17.417404174804688,
+      "learning_rate": 1.5015015015015017e-06,
+      "loss": 0.2934,
+      "step": 84900
+    },
+    {
+      "epoch": 254.98,
+      "grad_norm": 15.11248779296875,
+      "learning_rate": 1.5005005005005005e-06,
+      "loss": 0.3737,
+      "step": 84910
+    },
+    {
+      "epoch": 255.0,
+      "eval_accuracy": 0.8635,
+      "eval_loss": 0.5495705604553223,
+      "eval_runtime": 12.9391,
+      "eval_samples_per_second": 772.851,
+      "eval_steps_per_second": 3.091,
+      "step": 84915
+    },
+    {
+      "epoch": 255.02,
+      "grad_norm": 13.051982879638672,
+      "learning_rate": 1.4994994994994995e-06,
+      "loss": 0.3714,
+      "step": 84920
+    },
+    {
+      "epoch": 255.05,
+      "grad_norm": 13.010393142700195,
+      "learning_rate": 1.4984984984984986e-06,
+      "loss": 0.343,
+      "step": 84930
+    },
+    {
+      "epoch": 255.08,
+      "grad_norm": 15.865854263305664,
+      "learning_rate": 1.4974974974974976e-06,
+      "loss": 0.3016,
+      "step": 84940
+    },
+    {
+      "epoch": 255.11,
+      "grad_norm": 15.255988121032715,
+      "learning_rate": 1.4964964964964968e-06,
+      "loss": 0.3569,
+      "step": 84950
+    },
+    {
+      "epoch": 255.14,
+      "grad_norm": 19.419898986816406,
+      "learning_rate": 1.4954954954954956e-06,
+      "loss": 0.3494,
+      "step": 84960
+    },
+    {
+      "epoch": 255.17,
+      "grad_norm": 19.00016212463379,
+      "learning_rate": 1.4944944944944946e-06,
+      "loss": 0.3532,
+      "step": 84970
+    },
+    {
+      "epoch": 255.2,
+      "grad_norm": 12.750402450561523,
+      "learning_rate": 1.4934934934934937e-06,
+      "loss": 0.3026,
+      "step": 84980
+    },
+    {
+      "epoch": 255.23,
+      "grad_norm": 14.90976333618164,
+      "learning_rate": 1.4924924924924927e-06,
+      "loss": 0.3431,
+      "step": 84990
+    },
+    {
+      "epoch": 255.26,
+      "grad_norm": 14.477777481079102,
+      "learning_rate": 1.4914914914914915e-06,
+      "loss": 0.3443,
+      "step": 85000
+    },
+    {
+      "epoch": 255.29,
+      "grad_norm": 10.277405738830566,
+      "learning_rate": 1.4904904904904907e-06,
+      "loss": 0.3381,
+      "step": 85010
+    },
+    {
+      "epoch": 255.32,
+      "grad_norm": 21.07849884033203,
+      "learning_rate": 1.4894894894894895e-06,
+      "loss": 0.3293,
+      "step": 85020
+    },
+    {
+      "epoch": 255.35,
+      "grad_norm": 13.00082015991211,
+      "learning_rate": 1.4884884884884887e-06,
+      "loss": 0.3415,
+      "step": 85030
+    },
+    {
+      "epoch": 255.38,
+      "grad_norm": 10.960396766662598,
+      "learning_rate": 1.4874874874874876e-06,
+      "loss": 0.348,
+      "step": 85040
+    },
+    {
+      "epoch": 255.41,
+      "grad_norm": 21.659761428833008,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 0.3352,
+      "step": 85050
+    },
+    {
+      "epoch": 255.44,
+      "grad_norm": 24.280900955200195,
+      "learning_rate": 1.4854854854854856e-06,
+      "loss": 0.328,
+      "step": 85060
+    },
+    {
+      "epoch": 255.47,
+      "grad_norm": 14.043631553649902,
+      "learning_rate": 1.4844844844844846e-06,
+      "loss": 0.342,
+      "step": 85070
+    },
+    {
+      "epoch": 255.5,
+      "grad_norm": 13.094468116760254,
+      "learning_rate": 1.4834834834834836e-06,
+      "loss": 0.335,
+      "step": 85080
+    },
+    {
+      "epoch": 255.53,
+      "grad_norm": 15.323652267456055,
+      "learning_rate": 1.4824824824824826e-06,
+      "loss": 0.3046,
+      "step": 85090
+    },
+    {
+      "epoch": 255.56,
+      "grad_norm": 8.406963348388672,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.3074,
+      "step": 85100
+    },
+    {
+      "epoch": 255.59,
+      "grad_norm": 17.391469955444336,
+      "learning_rate": 1.4804804804804807e-06,
+      "loss": 0.366,
+      "step": 85110
+    },
+    {
+      "epoch": 255.62,
+      "grad_norm": 18.816524505615234,
+      "learning_rate": 1.4794794794794795e-06,
+      "loss": 0.3394,
+      "step": 85120
+    },
+    {
+      "epoch": 255.65,
+      "grad_norm": 15.168136596679688,
+      "learning_rate": 1.4784784784784787e-06,
+      "loss": 0.3125,
+      "step": 85130
+    },
+    {
+      "epoch": 255.68,
+      "grad_norm": 16.13576316833496,
+      "learning_rate": 1.4774774774774775e-06,
+      "loss": 0.3314,
+      "step": 85140
+    },
+    {
+      "epoch": 255.71,
+      "grad_norm": 14.324633598327637,
+      "learning_rate": 1.4764764764764765e-06,
+      "loss": 0.3622,
+      "step": 85150
+    },
+    {
+      "epoch": 255.74,
+      "grad_norm": 11.931784629821777,
+      "learning_rate": 1.4754754754754756e-06,
+      "loss": 0.3108,
+      "step": 85160
+    },
+    {
+      "epoch": 255.77,
+      "grad_norm": 10.952009201049805,
+      "learning_rate": 1.4744744744744746e-06,
+      "loss": 0.3536,
+      "step": 85170
+    },
+    {
+      "epoch": 255.8,
+      "grad_norm": 18.238758087158203,
+      "learning_rate": 1.4734734734734734e-06,
+      "loss": 0.3439,
+      "step": 85180
+    },
+    {
+      "epoch": 255.83,
+      "grad_norm": 12.951363563537598,
+      "learning_rate": 1.4724724724724726e-06,
+      "loss": 0.3184,
+      "step": 85190
+    },
+    {
+      "epoch": 255.86,
+      "grad_norm": 16.374713897705078,
+      "learning_rate": 1.4714714714714714e-06,
+      "loss": 0.3543,
+      "step": 85200
+    },
+    {
+      "epoch": 255.89,
+      "grad_norm": 17.104522705078125,
+      "learning_rate": 1.4704704704704707e-06,
+      "loss": 0.3705,
+      "step": 85210
+    },
+    {
+      "epoch": 255.92,
+      "grad_norm": 22.531490325927734,
+      "learning_rate": 1.4694694694694697e-06,
+      "loss": 0.3371,
+      "step": 85220
+    },
+    {
+      "epoch": 255.95,
+      "grad_norm": 14.134865760803223,
+      "learning_rate": 1.4684684684684685e-06,
+      "loss": 0.327,
+      "step": 85230
+    },
+    {
+      "epoch": 255.98,
+      "grad_norm": 21.47467803955078,
+      "learning_rate": 1.4674674674674677e-06,
+      "loss": 0.394,
+      "step": 85240
+    },
+    {
+      "epoch": 256.0,
+      "eval_accuracy": 0.8643,
+      "eval_loss": 0.5488690137863159,
+      "eval_runtime": 12.4689,
+      "eval_samples_per_second": 801.996,
+      "eval_steps_per_second": 3.208,
+      "step": 85248
+    },
+    {
+      "epoch": 256.01,
+      "grad_norm": 11.249968528747559,
+      "learning_rate": 1.4664664664664665e-06,
+      "loss": 0.3226,
+      "step": 85250
+    },
+    {
+      "epoch": 256.04,
+      "grad_norm": 14.24160099029541,
+      "learning_rate": 1.4654654654654657e-06,
+      "loss": 0.3115,
+      "step": 85260
+    },
+    {
+      "epoch": 256.07,
+      "grad_norm": 12.939754486083984,
+      "learning_rate": 1.4644644644644646e-06,
+      "loss": 0.3539,
+      "step": 85270
+    },
+    {
+      "epoch": 256.1,
+      "grad_norm": 10.749103546142578,
+      "learning_rate": 1.4634634634634638e-06,
+      "loss": 0.3261,
+      "step": 85280
+    },
+    {
+      "epoch": 256.13,
+      "grad_norm": 15.596113204956055,
+      "learning_rate": 1.4624624624624626e-06,
+      "loss": 0.3397,
+      "step": 85290
+    },
+    {
+      "epoch": 256.16,
+      "grad_norm": 13.873123168945312,
+      "learning_rate": 1.4614614614614616e-06,
+      "loss": 0.3181,
+      "step": 85300
+    },
+    {
+      "epoch": 256.19,
+      "grad_norm": 20.801929473876953,
+      "learning_rate": 1.4604604604604606e-06,
+      "loss": 0.3479,
+      "step": 85310
+    },
+    {
+      "epoch": 256.22,
+      "grad_norm": 12.698270797729492,
+      "learning_rate": 1.4594594594594596e-06,
+      "loss": 0.3484,
+      "step": 85320
+    },
+    {
+      "epoch": 256.25,
+      "grad_norm": 11.52580451965332,
+      "learning_rate": 1.4584584584584585e-06,
+      "loss": 0.3295,
+      "step": 85330
+    },
+    {
+      "epoch": 256.28,
+      "grad_norm": 19.337207794189453,
+      "learning_rate": 1.4574574574574577e-06,
+      "loss": 0.3135,
+      "step": 85340
+    },
+    {
+      "epoch": 256.31,
+      "grad_norm": 17.646787643432617,
+      "learning_rate": 1.4564564564564565e-06,
+      "loss": 0.3225,
+      "step": 85350
+    },
+    {
+      "epoch": 256.34,
+      "grad_norm": 14.842693328857422,
+      "learning_rate": 1.4554554554554557e-06,
+      "loss": 0.3221,
+      "step": 85360
+    },
+    {
+      "epoch": 256.37,
+      "grad_norm": 17.723609924316406,
+      "learning_rate": 1.4544544544544545e-06,
+      "loss": 0.3094,
+      "step": 85370
+    },
+    {
+      "epoch": 256.4,
+      "grad_norm": 14.534992218017578,
+      "learning_rate": 1.4534534534534535e-06,
+      "loss": 0.3112,
+      "step": 85380
+    },
+    {
+      "epoch": 256.43,
+      "grad_norm": 21.833599090576172,
+      "learning_rate": 1.4524524524524526e-06,
+      "loss": 0.331,
+      "step": 85390
+    },
+    {
+      "epoch": 256.46,
+      "grad_norm": 13.299799919128418,
+      "learning_rate": 1.4514514514514516e-06,
+      "loss": 0.3048,
+      "step": 85400
+    },
+    {
+      "epoch": 256.49,
+      "grad_norm": 11.56554126739502,
+      "learning_rate": 1.4504504504504504e-06,
+      "loss": 0.3304,
+      "step": 85410
+    },
+    {
+      "epoch": 256.52,
+      "grad_norm": 13.431511878967285,
+      "learning_rate": 1.4494494494494496e-06,
+      "loss": 0.3127,
+      "step": 85420
+    },
+    {
+      "epoch": 256.55,
+      "grad_norm": 15.66379165649414,
+      "learning_rate": 1.4484484484484484e-06,
+      "loss": 0.3707,
+      "step": 85430
+    },
+    {
+      "epoch": 256.58,
+      "grad_norm": 14.318967819213867,
+      "learning_rate": 1.4474474474474477e-06,
+      "loss": 0.3705,
+      "step": 85440
+    },
+    {
+      "epoch": 256.61,
+      "grad_norm": 30.597251892089844,
+      "learning_rate": 1.4464464464464465e-06,
+      "loss": 0.2989,
+      "step": 85450
+    },
+    {
+      "epoch": 256.64,
+      "grad_norm": 22.093067169189453,
+      "learning_rate": 1.4454454454454455e-06,
+      "loss": 0.3312,
+      "step": 85460
+    },
+    {
+      "epoch": 256.67,
+      "grad_norm": 23.069236755371094,
+      "learning_rate": 1.4444444444444445e-06,
+      "loss": 0.3563,
+      "step": 85470
+    },
+    {
+      "epoch": 256.7,
+      "grad_norm": 14.041702270507812,
+      "learning_rate": 1.4434434434434435e-06,
+      "loss": 0.3666,
+      "step": 85480
+    },
+    {
+      "epoch": 256.73,
+      "grad_norm": 17.68480110168457,
+      "learning_rate": 1.4424424424424427e-06,
+      "loss": 0.3272,
+      "step": 85490
+    },
+    {
+      "epoch": 256.76,
+      "grad_norm": 15.652909278869629,
+      "learning_rate": 1.4414414414414416e-06,
+      "loss": 0.3398,
+      "step": 85500
+    },
+    {
+      "epoch": 256.79,
+      "grad_norm": 12.619414329528809,
+      "learning_rate": 1.4404404404404408e-06,
+      "loss": 0.3303,
+      "step": 85510
+    },
+    {
+      "epoch": 256.82,
+      "grad_norm": 27.878454208374023,
+      "learning_rate": 1.4394394394394396e-06,
+      "loss": 0.303,
+      "step": 85520
+    },
+    {
+      "epoch": 256.85,
+      "grad_norm": 11.842232704162598,
+      "learning_rate": 1.4384384384384386e-06,
+      "loss": 0.3517,
+      "step": 85530
+    },
+    {
+      "epoch": 256.88,
+      "grad_norm": 9.285375595092773,
+      "learning_rate": 1.4374374374374376e-06,
+      "loss": 0.3311,
+      "step": 85540
+    },
+    {
+      "epoch": 256.91,
+      "grad_norm": 22.07879066467285,
+      "learning_rate": 1.4364364364364366e-06,
+      "loss": 0.3627,
+      "step": 85550
+    },
+    {
+      "epoch": 256.94,
+      "grad_norm": 18.763568878173828,
+      "learning_rate": 1.4354354354354354e-06,
+      "loss": 0.3136,
+      "step": 85560
+    },
+    {
+      "epoch": 256.97,
+      "grad_norm": 12.214506149291992,
+      "learning_rate": 1.4344344344344347e-06,
+      "loss": 0.3259,
+      "step": 85570
+    },
+    {
+      "epoch": 257.0,
+      "grad_norm": 18.285036087036133,
+      "learning_rate": 1.4334334334334335e-06,
+      "loss": 0.3394,
+      "step": 85580
+    },
+    {
+      "epoch": 257.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5513855218887329,
+      "eval_runtime": 12.8633,
+      "eval_samples_per_second": 777.404,
+      "eval_steps_per_second": 3.11,
+      "step": 85581
+    },
+    {
+      "epoch": 257.03,
+      "grad_norm": 23.281469345092773,
+      "learning_rate": 1.4324324324324327e-06,
+      "loss": 0.3076,
+      "step": 85590
+    },
+    {
+      "epoch": 257.06,
+      "grad_norm": 12.02914810180664,
+      "learning_rate": 1.4314314314314315e-06,
+      "loss": 0.2779,
+      "step": 85600
+    },
+    {
+      "epoch": 257.09,
+      "grad_norm": 17.252548217773438,
+      "learning_rate": 1.4304304304304305e-06,
+      "loss": 0.3496,
+      "step": 85610
+    },
+    {
+      "epoch": 257.12,
+      "grad_norm": 18.652265548706055,
+      "learning_rate": 1.4294294294294296e-06,
+      "loss": 0.3186,
+      "step": 85620
+    },
+    {
+      "epoch": 257.15,
+      "grad_norm": 35.09953308105469,
+      "learning_rate": 1.4284284284284286e-06,
+      "loss": 0.3957,
+      "step": 85630
+    },
+    {
+      "epoch": 257.18,
+      "grad_norm": 13.879796028137207,
+      "learning_rate": 1.4274274274274274e-06,
+      "loss": 0.2967,
+      "step": 85640
+    },
+    {
+      "epoch": 257.21,
+      "grad_norm": 16.474634170532227,
+      "learning_rate": 1.4264264264264266e-06,
+      "loss": 0.3154,
+      "step": 85650
+    },
+    {
+      "epoch": 257.24,
+      "grad_norm": 13.635601043701172,
+      "learning_rate": 1.4254254254254254e-06,
+      "loss": 0.3087,
+      "step": 85660
+    },
+    {
+      "epoch": 257.27,
+      "grad_norm": 13.972956657409668,
+      "learning_rate": 1.4244244244244247e-06,
+      "loss": 0.3153,
+      "step": 85670
+    },
+    {
+      "epoch": 257.3,
+      "grad_norm": 20.809417724609375,
+      "learning_rate": 1.4234234234234235e-06,
+      "loss": 0.3637,
+      "step": 85680
+    },
+    {
+      "epoch": 257.33,
+      "grad_norm": 11.593107223510742,
+      "learning_rate": 1.4224224224224225e-06,
+      "loss": 0.3436,
+      "step": 85690
+    },
+    {
+      "epoch": 257.36,
+      "grad_norm": 14.536643028259277,
+      "learning_rate": 1.4214214214214215e-06,
+      "loss": 0.3216,
+      "step": 85700
+    },
+    {
+      "epoch": 257.39,
+      "grad_norm": 15.46194076538086,
+      "learning_rate": 1.4204204204204205e-06,
+      "loss": 0.3573,
+      "step": 85710
+    },
+    {
+      "epoch": 257.42,
+      "grad_norm": 15.382064819335938,
+      "learning_rate": 1.4194194194194193e-06,
+      "loss": 0.3596,
+      "step": 85720
+    },
+    {
+      "epoch": 257.45,
+      "grad_norm": 16.776220321655273,
+      "learning_rate": 1.4184184184184185e-06,
+      "loss": 0.3236,
+      "step": 85730
+    },
+    {
+      "epoch": 257.48,
+      "grad_norm": 14.678390502929688,
+      "learning_rate": 1.4174174174174174e-06,
+      "loss": 0.3159,
+      "step": 85740
+    },
+    {
+      "epoch": 257.51,
+      "grad_norm": 15.51606273651123,
+      "learning_rate": 1.4164164164164166e-06,
+      "loss": 0.3392,
+      "step": 85750
+    },
+    {
+      "epoch": 257.54,
+      "grad_norm": 11.666996002197266,
+      "learning_rate": 1.4154154154154156e-06,
+      "loss": 0.3648,
+      "step": 85760
+    },
+    {
+      "epoch": 257.57,
+      "grad_norm": 16.869977951049805,
+      "learning_rate": 1.4144144144144144e-06,
+      "loss": 0.3306,
+      "step": 85770
+    },
+    {
+      "epoch": 257.6,
+      "grad_norm": 13.582656860351562,
+      "learning_rate": 1.4134134134134136e-06,
+      "loss": 0.3375,
+      "step": 85780
+    },
+    {
+      "epoch": 257.63,
+      "grad_norm": 16.567062377929688,
+      "learning_rate": 1.4124124124124124e-06,
+      "loss": 0.3373,
+      "step": 85790
+    },
+    {
+      "epoch": 257.66,
+      "grad_norm": 19.749563217163086,
+      "learning_rate": 1.4114114114114117e-06,
+      "loss": 0.2936,
+      "step": 85800
+    },
+    {
+      "epoch": 257.69,
+      "grad_norm": 16.821447372436523,
+      "learning_rate": 1.4104104104104105e-06,
+      "loss": 0.3666,
+      "step": 85810
+    },
+    {
+      "epoch": 257.72,
+      "grad_norm": 15.092323303222656,
+      "learning_rate": 1.4094094094094097e-06,
+      "loss": 0.3529,
+      "step": 85820
+    },
+    {
+      "epoch": 257.75,
+      "grad_norm": 12.867835998535156,
+      "learning_rate": 1.4084084084084085e-06,
+      "loss": 0.3297,
+      "step": 85830
+    },
+    {
+      "epoch": 257.78,
+      "grad_norm": 11.417281150817871,
+      "learning_rate": 1.4074074074074075e-06,
+      "loss": 0.3204,
+      "step": 85840
+    },
+    {
+      "epoch": 257.81,
+      "grad_norm": 18.96233558654785,
+      "learning_rate": 1.4064064064064066e-06,
+      "loss": 0.3247,
+      "step": 85850
+    },
+    {
+      "epoch": 257.84,
+      "grad_norm": 20.32769775390625,
+      "learning_rate": 1.4054054054054056e-06,
+      "loss": 0.344,
+      "step": 85860
+    },
+    {
+      "epoch": 257.87,
+      "grad_norm": 21.53523826599121,
+      "learning_rate": 1.4044044044044044e-06,
+      "loss": 0.3287,
+      "step": 85870
+    },
+    {
+      "epoch": 257.9,
+      "grad_norm": 18.957895278930664,
+      "learning_rate": 1.4034034034034036e-06,
+      "loss": 0.3493,
+      "step": 85880
+    },
+    {
+      "epoch": 257.93,
+      "grad_norm": 15.168993949890137,
+      "learning_rate": 1.4024024024024024e-06,
+      "loss": 0.3075,
+      "step": 85890
+    },
+    {
+      "epoch": 257.96,
+      "grad_norm": 13.454852104187012,
+      "learning_rate": 1.4014014014014016e-06,
+      "loss": 0.308,
+      "step": 85900
+    },
+    {
+      "epoch": 257.99,
+      "grad_norm": 12.91995906829834,
+      "learning_rate": 1.4004004004004005e-06,
+      "loss": 0.2963,
+      "step": 85910
+    },
+    {
+      "epoch": 258.0,
+      "eval_accuracy": 0.865,
+      "eval_loss": 0.5477760434150696,
+      "eval_runtime": 12.7516,
+      "eval_samples_per_second": 784.218,
+      "eval_steps_per_second": 3.137,
+      "step": 85914
+    },
+    {
+      "epoch": 258.02,
+      "grad_norm": 37.79804611206055,
+      "learning_rate": 1.3993993993993995e-06,
+      "loss": 0.2639,
+      "step": 85920
+    },
+    {
+      "epoch": 258.05,
+      "grad_norm": 17.77277374267578,
+      "learning_rate": 1.3983983983983985e-06,
+      "loss": 0.3572,
+      "step": 85930
+    },
+    {
+      "epoch": 258.08,
+      "grad_norm": 25.21531105041504,
+      "learning_rate": 1.3973973973973975e-06,
+      "loss": 0.3121,
+      "step": 85940
+    },
+    {
+      "epoch": 258.11,
+      "grad_norm": 14.828596115112305,
+      "learning_rate": 1.3963963963963963e-06,
+      "loss": 0.2912,
+      "step": 85950
+    },
+    {
+      "epoch": 258.14,
+      "grad_norm": 15.38594913482666,
+      "learning_rate": 1.3953953953953955e-06,
+      "loss": 0.3649,
+      "step": 85960
+    },
+    {
+      "epoch": 258.17,
+      "grad_norm": 17.138479232788086,
+      "learning_rate": 1.3943943943943944e-06,
+      "loss": 0.3622,
+      "step": 85970
+    },
+    {
+      "epoch": 258.2,
+      "grad_norm": 16.827590942382812,
+      "learning_rate": 1.3933933933933936e-06,
+      "loss": 0.375,
+      "step": 85980
+    },
+    {
+      "epoch": 258.23,
+      "grad_norm": 10.43038558959961,
+      "learning_rate": 1.3923923923923924e-06,
+      "loss": 0.3104,
+      "step": 85990
+    },
+    {
+      "epoch": 258.26,
+      "grad_norm": 15.718107223510742,
+      "learning_rate": 1.3913913913913914e-06,
+      "loss": 0.3889,
+      "step": 86000
+    },
+    {
+      "epoch": 258.29,
+      "grad_norm": 15.2860107421875,
+      "learning_rate": 1.3903903903903904e-06,
+      "loss": 0.3544,
+      "step": 86010
+    },
+    {
+      "epoch": 258.32,
+      "grad_norm": 18.010894775390625,
+      "learning_rate": 1.3893893893893894e-06,
+      "loss": 0.3234,
+      "step": 86020
+    },
+    {
+      "epoch": 258.35,
+      "grad_norm": 20.252845764160156,
+      "learning_rate": 1.3883883883883887e-06,
+      "loss": 0.3445,
+      "step": 86030
+    },
+    {
+      "epoch": 258.38,
+      "grad_norm": 16.194978713989258,
+      "learning_rate": 1.3873873873873875e-06,
+      "loss": 0.2902,
+      "step": 86040
+    },
+    {
+      "epoch": 258.41,
+      "grad_norm": 14.666404724121094,
+      "learning_rate": 1.3863863863863867e-06,
+      "loss": 0.3305,
+      "step": 86050
+    },
+    {
+      "epoch": 258.44,
+      "grad_norm": 15.970196723937988,
+      "learning_rate": 1.3853853853853855e-06,
+      "loss": 0.308,
+      "step": 86060
+    },
+    {
+      "epoch": 258.47,
+      "grad_norm": 17.528682708740234,
+      "learning_rate": 1.3843843843843845e-06,
+      "loss": 0.3727,
+      "step": 86070
+    },
+    {
+      "epoch": 258.5,
+      "grad_norm": 13.8905611038208,
+      "learning_rate": 1.3833833833833836e-06,
+      "loss": 0.3249,
+      "step": 86080
+    },
+    {
+      "epoch": 258.53,
+      "grad_norm": 13.141840934753418,
+      "learning_rate": 1.3823823823823826e-06,
+      "loss": 0.287,
+      "step": 86090
+    },
+    {
+      "epoch": 258.56,
+      "grad_norm": 13.95457935333252,
+      "learning_rate": 1.3813813813813814e-06,
+      "loss": 0.3403,
+      "step": 86100
+    },
+    {
+      "epoch": 258.59,
+      "grad_norm": 13.850568771362305,
+      "learning_rate": 1.3803803803803806e-06,
+      "loss": 0.3221,
+      "step": 86110
+    },
+    {
+      "epoch": 258.62,
+      "grad_norm": 9.761293411254883,
+      "learning_rate": 1.3793793793793794e-06,
+      "loss": 0.3065,
+      "step": 86120
+    },
+    {
+      "epoch": 258.65,
+      "grad_norm": 20.71546745300293,
+      "learning_rate": 1.3783783783783786e-06,
+      "loss": 0.3757,
+      "step": 86130
+    },
+    {
+      "epoch": 258.68,
+      "grad_norm": 15.858532905578613,
+      "learning_rate": 1.3773773773773775e-06,
+      "loss": 0.3811,
+      "step": 86140
+    },
+    {
+      "epoch": 258.71,
+      "grad_norm": 6.754410266876221,
+      "learning_rate": 1.3763763763763765e-06,
+      "loss": 0.3115,
+      "step": 86150
+    },
+    {
+      "epoch": 258.74,
+      "grad_norm": 24.534530639648438,
+      "learning_rate": 1.3753753753753755e-06,
+      "loss": 0.3492,
+      "step": 86160
+    },
+    {
+      "epoch": 258.77,
+      "grad_norm": 17.58386993408203,
+      "learning_rate": 1.3743743743743745e-06,
+      "loss": 0.3561,
+      "step": 86170
+    },
+    {
+      "epoch": 258.8,
+      "grad_norm": 15.564891815185547,
+      "learning_rate": 1.3733733733733733e-06,
+      "loss": 0.3528,
+      "step": 86180
+    },
+    {
+      "epoch": 258.83,
+      "grad_norm": 12.221661567687988,
+      "learning_rate": 1.3723723723723725e-06,
+      "loss": 0.335,
+      "step": 86190
+    },
+    {
+      "epoch": 258.86,
+      "grad_norm": 21.05199432373047,
+      "learning_rate": 1.3713713713713714e-06,
+      "loss": 0.3214,
+      "step": 86200
+    },
+    {
+      "epoch": 258.89,
+      "grad_norm": 22.06294822692871,
+      "learning_rate": 1.3703703703703706e-06,
+      "loss": 0.3332,
+      "step": 86210
+    },
+    {
+      "epoch": 258.92,
+      "grad_norm": 17.319807052612305,
+      "learning_rate": 1.3693693693693694e-06,
+      "loss": 0.3521,
+      "step": 86220
+    },
+    {
+      "epoch": 258.95,
+      "grad_norm": 17.26481056213379,
+      "learning_rate": 1.3683683683683684e-06,
+      "loss": 0.3402,
+      "step": 86230
+    },
+    {
+      "epoch": 258.98,
+      "grad_norm": 17.733253479003906,
+      "learning_rate": 1.3673673673673674e-06,
+      "loss": 0.3298,
+      "step": 86240
+    },
+    {
+      "epoch": 259.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.5483450889587402,
+      "eval_runtime": 12.9421,
+      "eval_samples_per_second": 772.671,
+      "eval_steps_per_second": 3.091,
+      "step": 86247
+    },
+    {
+      "epoch": 259.01,
+      "grad_norm": 16.119054794311523,
+      "learning_rate": 1.3663663663663664e-06,
+      "loss": 0.2832,
+      "step": 86250
+    },
+    {
+      "epoch": 259.04,
+      "grad_norm": 22.57053565979004,
+      "learning_rate": 1.3653653653653653e-06,
+      "loss": 0.3462,
+      "step": 86260
+    },
+    {
+      "epoch": 259.07,
+      "grad_norm": 26.16936492919922,
+      "learning_rate": 1.3643643643643645e-06,
+      "loss": 0.3522,
+      "step": 86270
+    },
+    {
+      "epoch": 259.1,
+      "grad_norm": 14.66234302520752,
+      "learning_rate": 1.3633633633633637e-06,
+      "loss": 0.3426,
+      "step": 86280
+    },
+    {
+      "epoch": 259.13,
+      "grad_norm": 13.921329498291016,
+      "learning_rate": 1.3623623623623625e-06,
+      "loss": 0.33,
+      "step": 86290
+    },
+    {
+      "epoch": 259.16,
+      "grad_norm": 15.972169876098633,
+      "learning_rate": 1.3613613613613615e-06,
+      "loss": 0.3605,
+      "step": 86300
+    },
+    {
+      "epoch": 259.19,
+      "grad_norm": 14.26353931427002,
+      "learning_rate": 1.3603603603603606e-06,
+      "loss": 0.3069,
+      "step": 86310
+    },
+    {
+      "epoch": 259.22,
+      "grad_norm": 14.106040954589844,
+      "learning_rate": 1.3593593593593596e-06,
+      "loss": 0.3184,
+      "step": 86320
+    },
+    {
+      "epoch": 259.25,
+      "grad_norm": 15.307622909545898,
+      "learning_rate": 1.3583583583583584e-06,
+      "loss": 0.3425,
+      "step": 86330
+    },
+    {
+      "epoch": 259.28,
+      "grad_norm": 9.22889518737793,
+      "learning_rate": 1.3573573573573576e-06,
+      "loss": 0.3287,
+      "step": 86340
+    },
+    {
+      "epoch": 259.31,
+      "grad_norm": 16.589460372924805,
+      "learning_rate": 1.3563563563563564e-06,
+      "loss": 0.3233,
+      "step": 86350
+    },
+    {
+      "epoch": 259.34,
+      "grad_norm": 16.608463287353516,
+      "learning_rate": 1.3553553553553556e-06,
+      "loss": 0.3431,
+      "step": 86360
+    },
+    {
+      "epoch": 259.37,
+      "grad_norm": 17.35122299194336,
+      "learning_rate": 1.3543543543543545e-06,
+      "loss": 0.2851,
+      "step": 86370
+    },
+    {
+      "epoch": 259.4,
+      "grad_norm": 18.37531852722168,
+      "learning_rate": 1.3533533533533535e-06,
+      "loss": 0.333,
+      "step": 86380
+    },
+    {
+      "epoch": 259.43,
+      "grad_norm": 19.0938777923584,
+      "learning_rate": 1.3523523523523525e-06,
+      "loss": 0.3971,
+      "step": 86390
+    },
+    {
+      "epoch": 259.46,
+      "grad_norm": 13.174006462097168,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.3188,
+      "step": 86400
+    },
+    {
+      "epoch": 259.49,
+      "grad_norm": 14.177721977233887,
+      "learning_rate": 1.3503503503503503e-06,
+      "loss": 0.3241,
+      "step": 86410
+    },
+    {
+      "epoch": 259.52,
+      "grad_norm": 20.978248596191406,
+      "learning_rate": 1.3493493493493495e-06,
+      "loss": 0.3394,
+      "step": 86420
+    },
+    {
+      "epoch": 259.55,
+      "grad_norm": 13.629049301147461,
+      "learning_rate": 1.3483483483483484e-06,
+      "loss": 0.3254,
+      "step": 86430
+    },
+    {
+      "epoch": 259.58,
+      "grad_norm": 18.68281364440918,
+      "learning_rate": 1.3473473473473476e-06,
+      "loss": 0.3244,
+      "step": 86440
+    },
+    {
+      "epoch": 259.61,
+      "grad_norm": 23.87210464477539,
+      "learning_rate": 1.3463463463463464e-06,
+      "loss": 0.3077,
+      "step": 86450
+    },
+    {
+      "epoch": 259.64,
+      "grad_norm": 10.668886184692383,
+      "learning_rate": 1.3453453453453454e-06,
+      "loss": 0.337,
+      "step": 86460
+    },
+    {
+      "epoch": 259.67,
+      "grad_norm": 11.048596382141113,
+      "learning_rate": 1.3443443443443444e-06,
+      "loss": 0.3399,
+      "step": 86470
+    },
+    {
+      "epoch": 259.7,
+      "grad_norm": 17.39781951904297,
+      "learning_rate": 1.3433433433433434e-06,
+      "loss": 0.3463,
+      "step": 86480
+    },
+    {
+      "epoch": 259.73,
+      "grad_norm": 21.962547302246094,
+      "learning_rate": 1.3423423423423422e-06,
+      "loss": 0.3229,
+      "step": 86490
+    },
+    {
+      "epoch": 259.76,
+      "grad_norm": 15.648326873779297,
+      "learning_rate": 1.3413413413413415e-06,
+      "loss": 0.3468,
+      "step": 86500
+    },
+    {
+      "epoch": 259.79,
+      "grad_norm": 12.10639476776123,
+      "learning_rate": 1.3403403403403403e-06,
+      "loss": 0.3163,
+      "step": 86510
+    },
+    {
+      "epoch": 259.82,
+      "grad_norm": 10.41342544555664,
+      "learning_rate": 1.3393393393393395e-06,
+      "loss": 0.2802,
+      "step": 86520
+    },
+    {
+      "epoch": 259.85,
+      "grad_norm": 15.65108585357666,
+      "learning_rate": 1.3383383383383383e-06,
+      "loss": 0.2845,
+      "step": 86530
+    },
+    {
+      "epoch": 259.88,
+      "grad_norm": 15.010077476501465,
+      "learning_rate": 1.3373373373373376e-06,
+      "loss": 0.2995,
+      "step": 86540
+    },
+    {
+      "epoch": 259.91,
+      "grad_norm": 12.219969749450684,
+      "learning_rate": 1.3363363363363366e-06,
+      "loss": 0.3058,
+      "step": 86550
+    },
+    {
+      "epoch": 259.94,
+      "grad_norm": 26.07997703552246,
+      "learning_rate": 1.3353353353353354e-06,
+      "loss": 0.3335,
+      "step": 86560
+    },
+    {
+      "epoch": 259.97,
+      "grad_norm": 20.135271072387695,
+      "learning_rate": 1.3343343343343346e-06,
+      "loss": 0.3526,
+      "step": 86570
+    },
+    {
+      "epoch": 260.0,
+      "grad_norm": 212.71788024902344,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.4112,
+      "step": 86580
+    },
+    {
+      "epoch": 260.0,
+      "eval_accuracy": 0.8637,
+      "eval_loss": 0.5491464138031006,
+      "eval_runtime": 12.7074,
+      "eval_samples_per_second": 786.943,
+      "eval_steps_per_second": 3.148,
+      "step": 86580
+    },
+    {
+      "epoch": 260.03,
+      "grad_norm": 34.932830810546875,
+      "learning_rate": 1.3323323323323326e-06,
+      "loss": 0.3019,
+      "step": 86590
+    },
+    {
+      "epoch": 260.06,
+      "grad_norm": 16.658897399902344,
+      "learning_rate": 1.3313313313313315e-06,
+      "loss": 0.2491,
+      "step": 86600
+    },
+    {
+      "epoch": 260.09,
+      "grad_norm": 13.306097030639648,
+      "learning_rate": 1.3303303303303305e-06,
+      "loss": 0.3167,
+      "step": 86610
+    },
+    {
+      "epoch": 260.12,
+      "grad_norm": 13.561806678771973,
+      "learning_rate": 1.3293293293293295e-06,
+      "loss": 0.3335,
+      "step": 86620
+    },
+    {
+      "epoch": 260.15,
+      "grad_norm": 20.90462875366211,
+      "learning_rate": 1.3283283283283285e-06,
+      "loss": 0.3552,
+      "step": 86630
+    },
+    {
+      "epoch": 260.18,
+      "grad_norm": 14.361353874206543,
+      "learning_rate": 1.3273273273273273e-06,
+      "loss": 0.3445,
+      "step": 86640
+    },
+    {
+      "epoch": 260.21,
+      "grad_norm": 15.049281120300293,
+      "learning_rate": 1.3263263263263265e-06,
+      "loss": 0.3129,
+      "step": 86650
+    },
+    {
+      "epoch": 260.24,
+      "grad_norm": 26.33973503112793,
+      "learning_rate": 1.3253253253253253e-06,
+      "loss": 0.2959,
+      "step": 86660
+    },
+    {
+      "epoch": 260.27,
+      "grad_norm": 22.85466957092285,
+      "learning_rate": 1.3243243243243246e-06,
+      "loss": 0.2792,
+      "step": 86670
+    },
+    {
+      "epoch": 260.3,
+      "grad_norm": 15.198979377746582,
+      "learning_rate": 1.3233233233233234e-06,
+      "loss": 0.3327,
+      "step": 86680
+    },
+    {
+      "epoch": 260.33,
+      "grad_norm": 12.365744590759277,
+      "learning_rate": 1.3223223223223224e-06,
+      "loss": 0.2975,
+      "step": 86690
+    },
+    {
+      "epoch": 260.36,
+      "grad_norm": 23.50188636779785,
+      "learning_rate": 1.3213213213213214e-06,
+      "loss": 0.3263,
+      "step": 86700
+    },
+    {
+      "epoch": 260.39,
+      "grad_norm": 13.288797378540039,
+      "learning_rate": 1.3203203203203204e-06,
+      "loss": 0.3781,
+      "step": 86710
+    },
+    {
+      "epoch": 260.42,
+      "grad_norm": 11.745049476623535,
+      "learning_rate": 1.3193193193193192e-06,
+      "loss": 0.3548,
+      "step": 86720
+    },
+    {
+      "epoch": 260.45,
+      "grad_norm": 18.491539001464844,
+      "learning_rate": 1.3183183183183185e-06,
+      "loss": 0.3089,
+      "step": 86730
+    },
+    {
+      "epoch": 260.48,
+      "grad_norm": 14.163981437683105,
+      "learning_rate": 1.3173173173173173e-06,
+      "loss": 0.3452,
+      "step": 86740
+    },
+    {
+      "epoch": 260.51,
+      "grad_norm": 19.954374313354492,
+      "learning_rate": 1.3163163163163165e-06,
+      "loss": 0.3701,
+      "step": 86750
+    },
+    {
+      "epoch": 260.54,
+      "grad_norm": 16.2740421295166,
+      "learning_rate": 1.3153153153153153e-06,
+      "loss": 0.3475,
+      "step": 86760
+    },
+    {
+      "epoch": 260.57,
+      "grad_norm": 21.159746170043945,
+      "learning_rate": 1.3143143143143143e-06,
+      "loss": 0.3272,
+      "step": 86770
+    },
+    {
+      "epoch": 260.6,
+      "grad_norm": 21.638195037841797,
+      "learning_rate": 1.3133133133133134e-06,
+      "loss": 0.37,
+      "step": 86780
+    },
+    {
+      "epoch": 260.63,
+      "grad_norm": 12.00246810913086,
+      "learning_rate": 1.3123123123123124e-06,
+      "loss": 0.3285,
+      "step": 86790
+    },
+    {
+      "epoch": 260.66,
+      "grad_norm": 13.608285903930664,
+      "learning_rate": 1.3113113113113112e-06,
+      "loss": 0.3266,
+      "step": 86800
+    },
+    {
+      "epoch": 260.69,
+      "grad_norm": 14.321954727172852,
+      "learning_rate": 1.3103103103103104e-06,
+      "loss": 0.3067,
+      "step": 86810
+    },
+    {
+      "epoch": 260.72,
+      "grad_norm": 17.36391258239746,
+      "learning_rate": 1.3093093093093096e-06,
+      "loss": 0.3176,
+      "step": 86820
+    },
+    {
+      "epoch": 260.75,
+      "grad_norm": 14.802751541137695,
+      "learning_rate": 1.3083083083083084e-06,
+      "loss": 0.3025,
+      "step": 86830
+    },
+    {
+      "epoch": 260.78,
+      "grad_norm": 18.525733947753906,
+      "learning_rate": 1.3073073073073075e-06,
+      "loss": 0.3367,
+      "step": 86840
+    },
+    {
+      "epoch": 260.81,
+      "grad_norm": 16.354604721069336,
+      "learning_rate": 1.3063063063063065e-06,
+      "loss": 0.3223,
+      "step": 86850
+    },
+    {
+      "epoch": 260.84,
+      "grad_norm": 13.113771438598633,
+      "learning_rate": 1.3053053053053055e-06,
+      "loss": 0.3257,
+      "step": 86860
+    },
+    {
+      "epoch": 260.87,
+      "grad_norm": 20.546592712402344,
+      "learning_rate": 1.3043043043043043e-06,
+      "loss": 0.3851,
+      "step": 86870
+    },
+    {
+      "epoch": 260.9,
+      "grad_norm": 14.350906372070312,
+      "learning_rate": 1.3033033033033035e-06,
+      "loss": 0.3114,
+      "step": 86880
+    },
+    {
+      "epoch": 260.93,
+      "grad_norm": 17.83962631225586,
+      "learning_rate": 1.3023023023023023e-06,
+      "loss": 0.3454,
+      "step": 86890
+    },
+    {
+      "epoch": 260.96,
+      "grad_norm": 17.095277786254883,
+      "learning_rate": 1.3013013013013016e-06,
+      "loss": 0.3726,
+      "step": 86900
+    },
+    {
+      "epoch": 260.99,
+      "grad_norm": 18.738985061645508,
+      "learning_rate": 1.3003003003003004e-06,
+      "loss": 0.3627,
+      "step": 86910
+    },
+    {
+      "epoch": 261.0,
+      "eval_accuracy": 0.8652,
+      "eval_loss": 0.5482470989227295,
+      "eval_runtime": 12.7423,
+      "eval_samples_per_second": 784.786,
+      "eval_steps_per_second": 3.139,
+      "step": 86913
+    },
+    {
+      "epoch": 261.02,
+      "grad_norm": 18.242393493652344,
+      "learning_rate": 1.2992992992992994e-06,
+      "loss": 0.346,
+      "step": 86920
+    },
+    {
+      "epoch": 261.05,
+      "grad_norm": 18.209074020385742,
+      "learning_rate": 1.2982982982982984e-06,
+      "loss": 0.3098,
+      "step": 86930
+    },
+    {
+      "epoch": 261.08,
+      "grad_norm": 16.10289192199707,
+      "learning_rate": 1.2972972972972974e-06,
+      "loss": 0.2841,
+      "step": 86940
+    },
+    {
+      "epoch": 261.11,
+      "grad_norm": 15.439473152160645,
+      "learning_rate": 1.2962962962962962e-06,
+      "loss": 0.3167,
+      "step": 86950
+    },
+    {
+      "epoch": 261.14,
+      "grad_norm": 13.523113250732422,
+      "learning_rate": 1.2952952952952955e-06,
+      "loss": 0.3457,
+      "step": 86960
+    },
+    {
+      "epoch": 261.17,
+      "grad_norm": 17.92822265625,
+      "learning_rate": 1.2942942942942943e-06,
+      "loss": 0.3894,
+      "step": 86970
+    },
+    {
+      "epoch": 261.2,
+      "grad_norm": 15.911784172058105,
+      "learning_rate": 1.2932932932932935e-06,
+      "loss": 0.3345,
+      "step": 86980
+    },
+    {
+      "epoch": 261.23,
+      "grad_norm": 11.576264381408691,
+      "learning_rate": 1.2922922922922923e-06,
+      "loss": 0.3391,
+      "step": 86990
+    },
+    {
+      "epoch": 261.26,
+      "grad_norm": 19.31475067138672,
+      "learning_rate": 1.2912912912912913e-06,
+      "loss": 0.3169,
+      "step": 87000
+    },
+    {
+      "epoch": 261.29,
+      "grad_norm": 12.393728256225586,
+      "learning_rate": 1.2902902902902904e-06,
+      "loss": 0.3308,
+      "step": 87010
+    },
+    {
+      "epoch": 261.32,
+      "grad_norm": 26.31382179260254,
+      "learning_rate": 1.2892892892892894e-06,
+      "loss": 0.2969,
+      "step": 87020
+    },
+    {
+      "epoch": 261.35,
+      "grad_norm": 17.476512908935547,
+      "learning_rate": 1.2882882882882882e-06,
+      "loss": 0.3593,
+      "step": 87030
+    },
+    {
+      "epoch": 261.38,
+      "grad_norm": 15.629095077514648,
+      "learning_rate": 1.2872872872872874e-06,
+      "loss": 0.3551,
+      "step": 87040
+    },
+    {
+      "epoch": 261.41,
+      "grad_norm": 14.06950569152832,
+      "learning_rate": 1.2862862862862862e-06,
+      "loss": 0.3438,
+      "step": 87050
+    },
+    {
+      "epoch": 261.44,
+      "grad_norm": 13.676071166992188,
+      "learning_rate": 1.2852852852852854e-06,
+      "loss": 0.3182,
+      "step": 87060
+    },
+    {
+      "epoch": 261.47,
+      "grad_norm": 16.516035079956055,
+      "learning_rate": 1.2842842842842843e-06,
+      "loss": 0.3746,
+      "step": 87070
+    },
+    {
+      "epoch": 261.5,
+      "grad_norm": 17.464630126953125,
+      "learning_rate": 1.2832832832832835e-06,
+      "loss": 0.3069,
+      "step": 87080
+    },
+    {
+      "epoch": 261.53,
+      "grad_norm": 19.894054412841797,
+      "learning_rate": 1.2822822822822825e-06,
+      "loss": 0.3037,
+      "step": 87090
+    },
+    {
+      "epoch": 261.56,
+      "grad_norm": 20.26616859436035,
+      "learning_rate": 1.2812812812812813e-06,
+      "loss": 0.2979,
+      "step": 87100
+    },
+    {
+      "epoch": 261.59,
+      "grad_norm": 30.044809341430664,
+      "learning_rate": 1.2802802802802805e-06,
+      "loss": 0.3562,
+      "step": 87110
+    },
+    {
+      "epoch": 261.62,
+      "grad_norm": 19.866947174072266,
+      "learning_rate": 1.2792792792792793e-06,
+      "loss": 0.4069,
+      "step": 87120
+    },
+    {
+      "epoch": 261.65,
+      "grad_norm": 15.52987289428711,
+      "learning_rate": 1.2782782782782786e-06,
+      "loss": 0.3244,
+      "step": 87130
+    },
+    {
+      "epoch": 261.68,
+      "grad_norm": 19.32136344909668,
+      "learning_rate": 1.2772772772772774e-06,
+      "loss": 0.329,
+      "step": 87140
+    },
+    {
+      "epoch": 261.71,
+      "grad_norm": 18.26842498779297,
+      "learning_rate": 1.2762762762762764e-06,
+      "loss": 0.3333,
+      "step": 87150
+    },
+    {
+      "epoch": 261.74,
+      "grad_norm": 13.469035148620605,
+      "learning_rate": 1.2752752752752754e-06,
+      "loss": 0.3209,
+      "step": 87160
+    },
+    {
+      "epoch": 261.77,
+      "grad_norm": 14.24643611907959,
+      "learning_rate": 1.2742742742742744e-06,
+      "loss": 0.3249,
+      "step": 87170
+    },
+    {
+      "epoch": 261.8,
+      "grad_norm": 19.474964141845703,
+      "learning_rate": 1.2732732732732732e-06,
+      "loss": 0.3185,
+      "step": 87180
+    },
+    {
+      "epoch": 261.83,
+      "grad_norm": 21.106361389160156,
+      "learning_rate": 1.2722722722722725e-06,
+      "loss": 0.3566,
+      "step": 87190
+    },
+    {
+      "epoch": 261.86,
+      "grad_norm": 17.282726287841797,
+      "learning_rate": 1.2712712712712713e-06,
+      "loss": 0.3744,
+      "step": 87200
+    },
+    {
+      "epoch": 261.89,
+      "grad_norm": 9.801228523254395,
+      "learning_rate": 1.2702702702702705e-06,
+      "loss": 0.3076,
+      "step": 87210
+    },
+    {
+      "epoch": 261.92,
+      "grad_norm": 22.351171493530273,
+      "learning_rate": 1.2692692692692693e-06,
+      "loss": 0.3609,
+      "step": 87220
+    },
+    {
+      "epoch": 261.95,
+      "grad_norm": 16.695247650146484,
+      "learning_rate": 1.2682682682682683e-06,
+      "loss": 0.3602,
+      "step": 87230
+    },
+    {
+      "epoch": 261.98,
+      "grad_norm": 12.652606010437012,
+      "learning_rate": 1.2672672672672674e-06,
+      "loss": 0.2939,
+      "step": 87240
+    },
+    {
+      "epoch": 262.0,
+      "eval_accuracy": 0.8649,
+      "eval_loss": 0.5480141043663025,
+      "eval_runtime": 12.8213,
+      "eval_samples_per_second": 779.951,
+      "eval_steps_per_second": 3.12,
+      "step": 87246
+    },
+    {
+      "epoch": 262.01,
+      "grad_norm": 17.066879272460938,
+      "learning_rate": 1.2662662662662664e-06,
+      "loss": 0.2429,
+      "step": 87250
+    },
+    {
+      "epoch": 262.04,
+      "grad_norm": 17.048500061035156,
+      "learning_rate": 1.2652652652652652e-06,
+      "loss": 0.3249,
+      "step": 87260
+    },
+    {
+      "epoch": 262.07,
+      "grad_norm": 19.921859741210938,
+      "learning_rate": 1.2642642642642644e-06,
+      "loss": 0.3287,
+      "step": 87270
+    },
+    {
+      "epoch": 262.1,
+      "grad_norm": 14.798046112060547,
+      "learning_rate": 1.2632632632632632e-06,
+      "loss": 0.2923,
+      "step": 87280
+    },
+    {
+      "epoch": 262.13,
+      "grad_norm": 19.927885055541992,
+      "learning_rate": 1.2622622622622624e-06,
+      "loss": 0.3468,
+      "step": 87290
+    },
+    {
+      "epoch": 262.16,
+      "grad_norm": 13.81885814666748,
+      "learning_rate": 1.2612612612612613e-06,
+      "loss": 0.3374,
+      "step": 87300
+    },
+    {
+      "epoch": 262.19,
+      "grad_norm": 18.351524353027344,
+      "learning_rate": 1.2602602602602605e-06,
+      "loss": 0.3047,
+      "step": 87310
+    },
+    {
+      "epoch": 262.22,
+      "grad_norm": 13.884198188781738,
+      "learning_rate": 1.2592592592592593e-06,
+      "loss": 0.3014,
+      "step": 87320
+    },
+    {
+      "epoch": 262.25,
+      "grad_norm": 17.980573654174805,
+      "learning_rate": 1.2582582582582583e-06,
+      "loss": 0.3188,
+      "step": 87330
+    },
+    {
+      "epoch": 262.28,
+      "grad_norm": 17.360620498657227,
+      "learning_rate": 1.2572572572572573e-06,
+      "loss": 0.3245,
+      "step": 87340
+    },
+    {
+      "epoch": 262.31,
+      "grad_norm": 19.209321975708008,
+      "learning_rate": 1.2562562562562563e-06,
+      "loss": 0.3108,
+      "step": 87350
+    },
+    {
+      "epoch": 262.34,
+      "grad_norm": 10.6846923828125,
+      "learning_rate": 1.2552552552552556e-06,
+      "loss": 0.2991,
+      "step": 87360
+    },
+    {
+      "epoch": 262.37,
+      "grad_norm": 16.49105453491211,
+      "learning_rate": 1.2542542542542544e-06,
+      "loss": 0.342,
+      "step": 87370
+    },
+    {
+      "epoch": 262.4,
+      "grad_norm": 24.448020935058594,
+      "learning_rate": 1.2532532532532534e-06,
+      "loss": 0.3473,
+      "step": 87380
+    },
+    {
+      "epoch": 262.43,
+      "grad_norm": 12.828956604003906,
+      "learning_rate": 1.2522522522522524e-06,
+      "loss": 0.2976,
+      "step": 87390
+    },
+    {
+      "epoch": 262.46,
+      "grad_norm": 17.232942581176758,
+      "learning_rate": 1.2512512512512514e-06,
+      "loss": 0.3429,
+      "step": 87400
+    },
+    {
+      "epoch": 262.49,
+      "grad_norm": 14.14938735961914,
+      "learning_rate": 1.2502502502502502e-06,
+      "loss": 0.3256,
+      "step": 87410
+    },
+    {
+      "epoch": 262.52,
+      "grad_norm": 13.68387222290039,
+      "learning_rate": 1.2492492492492493e-06,
+      "loss": 0.3233,
+      "step": 87420
+    },
+    {
+      "epoch": 262.55,
+      "grad_norm": 11.143026351928711,
+      "learning_rate": 1.2482482482482483e-06,
+      "loss": 0.2769,
+      "step": 87430
+    },
+    {
+      "epoch": 262.58,
+      "grad_norm": 12.524697303771973,
+      "learning_rate": 1.2472472472472473e-06,
+      "loss": 0.3603,
+      "step": 87440
+    },
+    {
+      "epoch": 262.61,
+      "grad_norm": 13.031341552734375,
+      "learning_rate": 1.2462462462462463e-06,
+      "loss": 0.2942,
+      "step": 87450
+    },
+    {
+      "epoch": 262.64,
+      "grad_norm": 13.474339485168457,
+      "learning_rate": 1.2452452452452453e-06,
+      "loss": 0.3373,
+      "step": 87460
+    },
+    {
+      "epoch": 262.67,
+      "grad_norm": 18.90024185180664,
+      "learning_rate": 1.2442442442442444e-06,
+      "loss": 0.3316,
+      "step": 87470
+    },
+    {
+      "epoch": 262.7,
+      "grad_norm": 17.423337936401367,
+      "learning_rate": 1.2432432432432434e-06,
+      "loss": 0.3095,
+      "step": 87480
+    },
+    {
+      "epoch": 262.73,
+      "grad_norm": 14.684900283813477,
+      "learning_rate": 1.2422422422422424e-06,
+      "loss": 0.3612,
+      "step": 87490
+    },
+    {
+      "epoch": 262.76,
+      "grad_norm": 16.887975692749023,
+      "learning_rate": 1.2412412412412414e-06,
+      "loss": 0.3291,
+      "step": 87500
+    },
+    {
+      "epoch": 262.79,
+      "grad_norm": 17.026918411254883,
+      "learning_rate": 1.2402402402402404e-06,
+      "loss": 0.3376,
+      "step": 87510
+    },
+    {
+      "epoch": 262.82,
+      "grad_norm": 14.101300239562988,
+      "learning_rate": 1.2392392392392394e-06,
+      "loss": 0.3166,
+      "step": 87520
+    },
+    {
+      "epoch": 262.85,
+      "grad_norm": 11.711414337158203,
+      "learning_rate": 1.2382382382382385e-06,
+      "loss": 0.3182,
+      "step": 87530
+    },
+    {
+      "epoch": 262.88,
+      "grad_norm": 13.427339553833008,
+      "learning_rate": 1.2372372372372375e-06,
+      "loss": 0.3204,
+      "step": 87540
+    },
+    {
+      "epoch": 262.91,
+      "grad_norm": 20.562679290771484,
+      "learning_rate": 1.2362362362362363e-06,
+      "loss": 0.3269,
+      "step": 87550
+    },
+    {
+      "epoch": 262.94,
+      "grad_norm": 18.792972564697266,
+      "learning_rate": 1.2352352352352353e-06,
+      "loss": 0.3495,
+      "step": 87560
+    },
+    {
+      "epoch": 262.97,
+      "grad_norm": 28.865753173828125,
+      "learning_rate": 1.2342342342342343e-06,
+      "loss": 0.2827,
+      "step": 87570
+    },
+    {
+      "epoch": 263.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5500628352165222,
+      "eval_runtime": 13.0135,
+      "eval_samples_per_second": 768.432,
+      "eval_steps_per_second": 3.074,
+      "step": 87579
+    },
+    {
+      "epoch": 263.0,
+      "grad_norm": 20.31890296936035,
+      "learning_rate": 1.2332332332332333e-06,
+      "loss": 0.2954,
+      "step": 87580
+    },
+    {
+      "epoch": 263.03,
+      "grad_norm": 14.902915000915527,
+      "learning_rate": 1.2322322322322324e-06,
+      "loss": 0.294,
+      "step": 87590
+    },
+    {
+      "epoch": 263.06,
+      "grad_norm": 15.590399742126465,
+      "learning_rate": 1.2312312312312314e-06,
+      "loss": 0.3015,
+      "step": 87600
+    },
+    {
+      "epoch": 263.09,
+      "grad_norm": 14.715773582458496,
+      "learning_rate": 1.2302302302302304e-06,
+      "loss": 0.3192,
+      "step": 87610
+    },
+    {
+      "epoch": 263.12,
+      "grad_norm": 11.64497184753418,
+      "learning_rate": 1.2292292292292294e-06,
+      "loss": 0.3237,
+      "step": 87620
+    },
+    {
+      "epoch": 263.15,
+      "grad_norm": 21.231760025024414,
+      "learning_rate": 1.2282282282282282e-06,
+      "loss": 0.328,
+      "step": 87630
+    },
+    {
+      "epoch": 263.18,
+      "grad_norm": 14.169036865234375,
+      "learning_rate": 1.2272272272272272e-06,
+      "loss": 0.2998,
+      "step": 87640
+    },
+    {
+      "epoch": 263.21,
+      "grad_norm": 18.05064582824707,
+      "learning_rate": 1.2262262262262263e-06,
+      "loss": 0.3461,
+      "step": 87650
+    },
+    {
+      "epoch": 263.24,
+      "grad_norm": 17.682573318481445,
+      "learning_rate": 1.2252252252252253e-06,
+      "loss": 0.3333,
+      "step": 87660
+    },
+    {
+      "epoch": 263.27,
+      "grad_norm": 23.082866668701172,
+      "learning_rate": 1.2242242242242243e-06,
+      "loss": 0.2584,
+      "step": 87670
+    },
+    {
+      "epoch": 263.3,
+      "grad_norm": 16.512319564819336,
+      "learning_rate": 1.2232232232232233e-06,
+      "loss": 0.3144,
+      "step": 87680
+    },
+    {
+      "epoch": 263.33,
+      "grad_norm": 39.2318115234375,
+      "learning_rate": 1.2222222222222223e-06,
+      "loss": 0.361,
+      "step": 87690
+    },
+    {
+      "epoch": 263.36,
+      "grad_norm": 18.54353141784668,
+      "learning_rate": 1.2212212212212213e-06,
+      "loss": 0.3079,
+      "step": 87700
+    },
+    {
+      "epoch": 263.39,
+      "grad_norm": 14.141229629516602,
+      "learning_rate": 1.2202202202202204e-06,
+      "loss": 0.3241,
+      "step": 87710
+    },
+    {
+      "epoch": 263.42,
+      "grad_norm": 19.690078735351562,
+      "learning_rate": 1.2192192192192192e-06,
+      "loss": 0.2815,
+      "step": 87720
+    },
+    {
+      "epoch": 263.45,
+      "grad_norm": 16.804244995117188,
+      "learning_rate": 1.2182182182182182e-06,
+      "loss": 0.3228,
+      "step": 87730
+    },
+    {
+      "epoch": 263.48,
+      "grad_norm": 11.556256294250488,
+      "learning_rate": 1.2172172172172172e-06,
+      "loss": 0.3621,
+      "step": 87740
+    },
+    {
+      "epoch": 263.51,
+      "grad_norm": 13.593873023986816,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 0.3182,
+      "step": 87750
+    },
+    {
+      "epoch": 263.54,
+      "grad_norm": 18.875795364379883,
+      "learning_rate": 1.2152152152152155e-06,
+      "loss": 0.2852,
+      "step": 87760
+    },
+    {
+      "epoch": 263.57,
+      "grad_norm": 9.71463394165039,
+      "learning_rate": 1.2142142142142143e-06,
+      "loss": 0.3012,
+      "step": 87770
+    },
+    {
+      "epoch": 263.6,
+      "grad_norm": 9.229799270629883,
+      "learning_rate": 1.2132132132132133e-06,
+      "loss": 0.3408,
+      "step": 87780
+    },
+    {
+      "epoch": 263.63,
+      "grad_norm": 12.624789237976074,
+      "learning_rate": 1.2122122122122123e-06,
+      "loss": 0.2936,
+      "step": 87790
+    },
+    {
+      "epoch": 263.66,
+      "grad_norm": 15.971983909606934,
+      "learning_rate": 1.2112112112112113e-06,
+      "loss": 0.2932,
+      "step": 87800
+    },
+    {
+      "epoch": 263.69,
+      "grad_norm": 15.99201774597168,
+      "learning_rate": 1.2102102102102103e-06,
+      "loss": 0.3404,
+      "step": 87810
+    },
+    {
+      "epoch": 263.72,
+      "grad_norm": 20.064159393310547,
+      "learning_rate": 1.2092092092092094e-06,
+      "loss": 0.3139,
+      "step": 87820
+    },
+    {
+      "epoch": 263.75,
+      "grad_norm": 14.729034423828125,
+      "learning_rate": 1.2082082082082084e-06,
+      "loss": 0.3206,
+      "step": 87830
+    },
+    {
+      "epoch": 263.78,
+      "grad_norm": 19.855880737304688,
+      "learning_rate": 1.2072072072072074e-06,
+      "loss": 0.3598,
+      "step": 87840
+    },
+    {
+      "epoch": 263.81,
+      "grad_norm": 22.034469604492188,
+      "learning_rate": 1.2062062062062064e-06,
+      "loss": 0.2996,
+      "step": 87850
+    },
+    {
+      "epoch": 263.84,
+      "grad_norm": 15.21556282043457,
+      "learning_rate": 1.2052052052052052e-06,
+      "loss": 0.3465,
+      "step": 87860
+    },
+    {
+      "epoch": 263.87,
+      "grad_norm": 17.12798500061035,
+      "learning_rate": 1.2042042042042042e-06,
+      "loss": 0.3365,
+      "step": 87870
+    },
+    {
+      "epoch": 263.9,
+      "grad_norm": 14.423870086669922,
+      "learning_rate": 1.2032032032032033e-06,
+      "loss": 0.2876,
+      "step": 87880
+    },
+    {
+      "epoch": 263.93,
+      "grad_norm": 15.651647567749023,
+      "learning_rate": 1.2022022022022023e-06,
+      "loss": 0.3922,
+      "step": 87890
+    },
+    {
+      "epoch": 263.96,
+      "grad_norm": 19.778121948242188,
+      "learning_rate": 1.2012012012012013e-06,
+      "loss": 0.3247,
+      "step": 87900
+    },
+    {
+      "epoch": 263.99,
+      "grad_norm": 17.766372680664062,
+      "learning_rate": 1.2002002002002003e-06,
+      "loss": 0.3001,
+      "step": 87910
+    },
+    {
+      "epoch": 264.0,
+      "eval_accuracy": 0.8659,
+      "eval_loss": 0.5491458773612976,
+      "eval_runtime": 12.4885,
+      "eval_samples_per_second": 800.734,
+      "eval_steps_per_second": 3.203,
+      "step": 87912
+    },
+    {
+      "epoch": 264.02,
+      "grad_norm": 11.736800193786621,
+      "learning_rate": 1.1991991991991993e-06,
+      "loss": 0.3162,
+      "step": 87920
+    },
+    {
+      "epoch": 264.05,
+      "grad_norm": 17.539796829223633,
+      "learning_rate": 1.1981981981981983e-06,
+      "loss": 0.2889,
+      "step": 87930
+    },
+    {
+      "epoch": 264.08,
+      "grad_norm": 16.699207305908203,
+      "learning_rate": 1.1971971971971974e-06,
+      "loss": 0.3441,
+      "step": 87940
+    },
+    {
+      "epoch": 264.11,
+      "grad_norm": 13.979104995727539,
+      "learning_rate": 1.1961961961961962e-06,
+      "loss": 0.312,
+      "step": 87950
+    },
+    {
+      "epoch": 264.14,
+      "grad_norm": 16.83622169494629,
+      "learning_rate": 1.1951951951951952e-06,
+      "loss": 0.3416,
+      "step": 87960
+    },
+    {
+      "epoch": 264.17,
+      "grad_norm": 12.434218406677246,
+      "learning_rate": 1.1941941941941942e-06,
+      "loss": 0.3197,
+      "step": 87970
+    },
+    {
+      "epoch": 264.2,
+      "grad_norm": 26.15000343322754,
+      "learning_rate": 1.1931931931931932e-06,
+      "loss": 0.3232,
+      "step": 87980
+    },
+    {
+      "epoch": 264.23,
+      "grad_norm": 11.62374496459961,
+      "learning_rate": 1.1921921921921922e-06,
+      "loss": 0.2982,
+      "step": 87990
+    },
+    {
+      "epoch": 264.26,
+      "grad_norm": 17.26070213317871,
+      "learning_rate": 1.1911911911911913e-06,
+      "loss": 0.3152,
+      "step": 88000
+    },
+    {
+      "epoch": 264.29,
+      "grad_norm": 19.604610443115234,
+      "learning_rate": 1.1901901901901903e-06,
+      "loss": 0.3258,
+      "step": 88010
+    },
+    {
+      "epoch": 264.32,
+      "grad_norm": 14.040647506713867,
+      "learning_rate": 1.1891891891891893e-06,
+      "loss": 0.2817,
+      "step": 88020
+    },
+    {
+      "epoch": 264.35,
+      "grad_norm": 14.941747665405273,
+      "learning_rate": 1.1881881881881883e-06,
+      "loss": 0.3353,
+      "step": 88030
+    },
+    {
+      "epoch": 264.38,
+      "grad_norm": 22.88564109802246,
+      "learning_rate": 1.1871871871871873e-06,
+      "loss": 0.3424,
+      "step": 88040
+    },
+    {
+      "epoch": 264.41,
+      "grad_norm": 10.367718696594238,
+      "learning_rate": 1.1861861861861864e-06,
+      "loss": 0.3463,
+      "step": 88050
+    },
+    {
+      "epoch": 264.44,
+      "grad_norm": 20.72346305847168,
+      "learning_rate": 1.1851851851851854e-06,
+      "loss": 0.2996,
+      "step": 88060
+    },
+    {
+      "epoch": 264.47,
+      "grad_norm": 15.990628242492676,
+      "learning_rate": 1.1841841841841844e-06,
+      "loss": 0.34,
+      "step": 88070
+    },
+    {
+      "epoch": 264.5,
+      "grad_norm": 20.239517211914062,
+      "learning_rate": 1.1831831831831834e-06,
+      "loss": 0.324,
+      "step": 88080
+    },
+    {
+      "epoch": 264.53,
+      "grad_norm": 29.40633773803711,
+      "learning_rate": 1.1821821821821822e-06,
+      "loss": 0.3597,
+      "step": 88090
+    },
+    {
+      "epoch": 264.56,
+      "grad_norm": 21.207571029663086,
+      "learning_rate": 1.1811811811811812e-06,
+      "loss": 0.3118,
+      "step": 88100
+    },
+    {
+      "epoch": 264.59,
+      "grad_norm": 13.606197357177734,
+      "learning_rate": 1.1801801801801803e-06,
+      "loss": 0.335,
+      "step": 88110
+    },
+    {
+      "epoch": 264.62,
+      "grad_norm": 17.49561309814453,
+      "learning_rate": 1.1791791791791793e-06,
+      "loss": 0.3734,
+      "step": 88120
+    },
+    {
+      "epoch": 264.65,
+      "grad_norm": 15.865230560302734,
+      "learning_rate": 1.1781781781781783e-06,
+      "loss": 0.3332,
+      "step": 88130
+    },
+    {
+      "epoch": 264.68,
+      "grad_norm": 14.792943000793457,
+      "learning_rate": 1.1771771771771773e-06,
+      "loss": 0.3252,
+      "step": 88140
+    },
+    {
+      "epoch": 264.71,
+      "grad_norm": 13.440323829650879,
+      "learning_rate": 1.1761761761761763e-06,
+      "loss": 0.3869,
+      "step": 88150
+    },
+    {
+      "epoch": 264.74,
+      "grad_norm": 22.474925994873047,
+      "learning_rate": 1.1751751751751753e-06,
+      "loss": 0.311,
+      "step": 88160
+    },
+    {
+      "epoch": 264.77,
+      "grad_norm": 28.534141540527344,
+      "learning_rate": 1.1741741741741742e-06,
+      "loss": 0.31,
+      "step": 88170
+    },
+    {
+      "epoch": 264.8,
+      "grad_norm": 18.39545249938965,
+      "learning_rate": 1.1731731731731732e-06,
+      "loss": 0.3402,
+      "step": 88180
+    },
+    {
+      "epoch": 264.83,
+      "grad_norm": 16.371299743652344,
+      "learning_rate": 1.1721721721721722e-06,
+      "loss": 0.3315,
+      "step": 88190
+    },
+    {
+      "epoch": 264.86,
+      "grad_norm": 10.318588256835938,
+      "learning_rate": 1.1711711711711712e-06,
+      "loss": 0.3141,
+      "step": 88200
+    },
+    {
+      "epoch": 264.89,
+      "grad_norm": 10.11978816986084,
+      "learning_rate": 1.1701701701701702e-06,
+      "loss": 0.3028,
+      "step": 88210
+    },
+    {
+      "epoch": 264.92,
+      "grad_norm": 12.660517692565918,
+      "learning_rate": 1.1691691691691692e-06,
+      "loss": 0.2867,
+      "step": 88220
+    },
+    {
+      "epoch": 264.95,
+      "grad_norm": 16.254793167114258,
+      "learning_rate": 1.1681681681681683e-06,
+      "loss": 0.285,
+      "step": 88230
+    },
+    {
+      "epoch": 264.98,
+      "grad_norm": 11.154897689819336,
+      "learning_rate": 1.1671671671671673e-06,
+      "loss": 0.3106,
+      "step": 88240
+    },
+    {
+      "epoch": 265.0,
+      "eval_accuracy": 0.8662,
+      "eval_loss": 0.5485965609550476,
+      "eval_runtime": 12.6121,
+      "eval_samples_per_second": 792.891,
+      "eval_steps_per_second": 3.172,
+      "step": 88245
+    },
+    {
+      "epoch": 265.02,
+      "grad_norm": 16.29247283935547,
+      "learning_rate": 1.1661661661661663e-06,
+      "loss": 0.3415,
+      "step": 88250
+    },
+    {
+      "epoch": 265.05,
+      "grad_norm": 9.637909889221191,
+      "learning_rate": 1.1651651651651651e-06,
+      "loss": 0.3169,
+      "step": 88260
+    },
+    {
+      "epoch": 265.08,
+      "grad_norm": 19.214853286743164,
+      "learning_rate": 1.1641641641641641e-06,
+      "loss": 0.3336,
+      "step": 88270
+    },
+    {
+      "epoch": 265.11,
+      "grad_norm": 13.82374382019043,
+      "learning_rate": 1.1631631631631631e-06,
+      "loss": 0.272,
+      "step": 88280
+    },
+    {
+      "epoch": 265.14,
+      "grad_norm": 34.17050552368164,
+      "learning_rate": 1.1621621621621624e-06,
+      "loss": 0.3596,
+      "step": 88290
+    },
+    {
+      "epoch": 265.17,
+      "grad_norm": 13.722670555114746,
+      "learning_rate": 1.1611611611611614e-06,
+      "loss": 0.3339,
+      "step": 88300
+    },
+    {
+      "epoch": 265.2,
+      "grad_norm": 15.15519905090332,
+      "learning_rate": 1.1601601601601604e-06,
+      "loss": 0.357,
+      "step": 88310
+    },
+    {
+      "epoch": 265.23,
+      "grad_norm": 17.565900802612305,
+      "learning_rate": 1.1591591591591592e-06,
+      "loss": 0.3415,
+      "step": 88320
+    },
+    {
+      "epoch": 265.26,
+      "grad_norm": 22.63159942626953,
+      "learning_rate": 1.1581581581581582e-06,
+      "loss": 0.3205,
+      "step": 88330
+    },
+    {
+      "epoch": 265.29,
+      "grad_norm": 11.41580867767334,
+      "learning_rate": 1.1571571571571573e-06,
+      "loss": 0.3128,
+      "step": 88340
+    },
+    {
+      "epoch": 265.32,
+      "grad_norm": 13.753207206726074,
+      "learning_rate": 1.1561561561561563e-06,
+      "loss": 0.2774,
+      "step": 88350
+    },
+    {
+      "epoch": 265.35,
+      "grad_norm": 14.797466278076172,
+      "learning_rate": 1.1551551551551553e-06,
+      "loss": 0.2933,
+      "step": 88360
+    },
+    {
+      "epoch": 265.38,
+      "grad_norm": 13.945099830627441,
+      "learning_rate": 1.1541541541541543e-06,
+      "loss": 0.3395,
+      "step": 88370
+    },
+    {
+      "epoch": 265.41,
+      "grad_norm": 14.228446006774902,
+      "learning_rate": 1.1531531531531533e-06,
+      "loss": 0.3368,
+      "step": 88380
+    },
+    {
+      "epoch": 265.44,
+      "grad_norm": 11.859404563903809,
+      "learning_rate": 1.1521521521521523e-06,
+      "loss": 0.346,
+      "step": 88390
+    },
+    {
+      "epoch": 265.47,
+      "grad_norm": 29.687620162963867,
+      "learning_rate": 1.1511511511511512e-06,
+      "loss": 0.3487,
+      "step": 88400
+    },
+    {
+      "epoch": 265.5,
+      "grad_norm": 24.21990966796875,
+      "learning_rate": 1.1501501501501502e-06,
+      "loss": 0.332,
+      "step": 88410
+    },
+    {
+      "epoch": 265.53,
+      "grad_norm": 16.470687866210938,
+      "learning_rate": 1.1491491491491492e-06,
+      "loss": 0.2911,
+      "step": 88420
+    },
+    {
+      "epoch": 265.56,
+      "grad_norm": 15.00847053527832,
+      "learning_rate": 1.1481481481481482e-06,
+      "loss": 0.3298,
+      "step": 88430
+    },
+    {
+      "epoch": 265.59,
+      "grad_norm": 12.42449951171875,
+      "learning_rate": 1.1471471471471472e-06,
+      "loss": 0.3453,
+      "step": 88440
+    },
+    {
+      "epoch": 265.62,
+      "grad_norm": 15.026987075805664,
+      "learning_rate": 1.1461461461461462e-06,
+      "loss": 0.3158,
+      "step": 88450
+    },
+    {
+      "epoch": 265.65,
+      "grad_norm": 23.484373092651367,
+      "learning_rate": 1.1451451451451453e-06,
+      "loss": 0.3267,
+      "step": 88460
+    },
+    {
+      "epoch": 265.68,
+      "grad_norm": 19.459096908569336,
+      "learning_rate": 1.1441441441441443e-06,
+      "loss": 0.3392,
+      "step": 88470
+    },
+    {
+      "epoch": 265.71,
+      "grad_norm": 17.21712303161621,
+      "learning_rate": 1.1431431431431433e-06,
+      "loss": 0.2923,
+      "step": 88480
+    },
+    {
+      "epoch": 265.74,
+      "grad_norm": 14.88533878326416,
+      "learning_rate": 1.142142142142142e-06,
+      "loss": 0.306,
+      "step": 88490
+    },
+    {
+      "epoch": 265.77,
+      "grad_norm": 15.648037910461426,
+      "learning_rate": 1.1411411411411411e-06,
+      "loss": 0.3213,
+      "step": 88500
+    },
+    {
+      "epoch": 265.8,
+      "grad_norm": 13.435467720031738,
+      "learning_rate": 1.1401401401401401e-06,
+      "loss": 0.3686,
+      "step": 88510
+    },
+    {
+      "epoch": 265.83,
+      "grad_norm": 11.270416259765625,
+      "learning_rate": 1.1391391391391392e-06,
+      "loss": 0.3701,
+      "step": 88520
+    },
+    {
+      "epoch": 265.86,
+      "grad_norm": 15.472171783447266,
+      "learning_rate": 1.1381381381381382e-06,
+      "loss": 0.3461,
+      "step": 88530
+    },
+    {
+      "epoch": 265.89,
+      "grad_norm": 17.703062057495117,
+      "learning_rate": 1.1371371371371372e-06,
+      "loss": 0.328,
+      "step": 88540
+    },
+    {
+      "epoch": 265.92,
+      "grad_norm": 37.11611557006836,
+      "learning_rate": 1.1361361361361362e-06,
+      "loss": 0.3302,
+      "step": 88550
+    },
+    {
+      "epoch": 265.95,
+      "grad_norm": 13.46673583984375,
+      "learning_rate": 1.1351351351351352e-06,
+      "loss": 0.3259,
+      "step": 88560
+    },
+    {
+      "epoch": 265.98,
+      "grad_norm": 18.108558654785156,
+      "learning_rate": 1.1341341341341343e-06,
+      "loss": 0.3416,
+      "step": 88570
+    },
+    {
+      "epoch": 266.0,
+      "eval_accuracy": 0.8641,
+      "eval_loss": 0.5509369373321533,
+      "eval_runtime": 12.4024,
+      "eval_samples_per_second": 806.298,
+      "eval_steps_per_second": 3.225,
+      "step": 88578
+    },
+    {
+      "epoch": 266.01,
+      "grad_norm": 13.582796096801758,
+      "learning_rate": 1.1331331331331333e-06,
+      "loss": 0.4009,
+      "step": 88580
+    },
+    {
+      "epoch": 266.04,
+      "grad_norm": 14.023848533630371,
+      "learning_rate": 1.1321321321321323e-06,
+      "loss": 0.3297,
+      "step": 88590
+    },
+    {
+      "epoch": 266.07,
+      "grad_norm": 12.389493942260742,
+      "learning_rate": 1.1311311311311313e-06,
+      "loss": 0.3283,
+      "step": 88600
+    },
+    {
+      "epoch": 266.1,
+      "grad_norm": 12.123244285583496,
+      "learning_rate": 1.1301301301301303e-06,
+      "loss": 0.3057,
+      "step": 88610
+    },
+    {
+      "epoch": 266.13,
+      "grad_norm": 14.076654434204102,
+      "learning_rate": 1.1291291291291293e-06,
+      "loss": 0.3277,
+      "step": 88620
+    },
+    {
+      "epoch": 266.16,
+      "grad_norm": 22.83381462097168,
+      "learning_rate": 1.1281281281281281e-06,
+      "loss": 0.3317,
+      "step": 88630
+    },
+    {
+      "epoch": 266.19,
+      "grad_norm": 23.250593185424805,
+      "learning_rate": 1.1271271271271272e-06,
+      "loss": 0.3612,
+      "step": 88640
+    },
+    {
+      "epoch": 266.22,
+      "grad_norm": 11.257255554199219,
+      "learning_rate": 1.1261261261261262e-06,
+      "loss": 0.3452,
+      "step": 88650
+    },
+    {
+      "epoch": 266.25,
+      "grad_norm": 18.001585006713867,
+      "learning_rate": 1.1251251251251252e-06,
+      "loss": 0.3024,
+      "step": 88660
+    },
+    {
+      "epoch": 266.28,
+      "grad_norm": 18.87581443786621,
+      "learning_rate": 1.1241241241241242e-06,
+      "loss": 0.3687,
+      "step": 88670
+    },
+    {
+      "epoch": 266.31,
+      "grad_norm": 19.574838638305664,
+      "learning_rate": 1.1231231231231232e-06,
+      "loss": 0.3301,
+      "step": 88680
+    },
+    {
+      "epoch": 266.34,
+      "grad_norm": 19.74444580078125,
+      "learning_rate": 1.1221221221221223e-06,
+      "loss": 0.3345,
+      "step": 88690
+    },
+    {
+      "epoch": 266.37,
+      "grad_norm": 30.848941802978516,
+      "learning_rate": 1.1211211211211213e-06,
+      "loss": 0.3589,
+      "step": 88700
+    },
+    {
+      "epoch": 266.4,
+      "grad_norm": 12.555103302001953,
+      "learning_rate": 1.1201201201201203e-06,
+      "loss": 0.3285,
+      "step": 88710
+    },
+    {
+      "epoch": 266.43,
+      "grad_norm": 19.571474075317383,
+      "learning_rate": 1.119119119119119e-06,
+      "loss": 0.294,
+      "step": 88720
+    },
+    {
+      "epoch": 266.46,
+      "grad_norm": 15.353864669799805,
+      "learning_rate": 1.1181181181181181e-06,
+      "loss": 0.3583,
+      "step": 88730
+    },
+    {
+      "epoch": 266.49,
+      "grad_norm": 15.16854190826416,
+      "learning_rate": 1.1171171171171171e-06,
+      "loss": 0.2897,
+      "step": 88740
+    },
+    {
+      "epoch": 266.52,
+      "grad_norm": 15.02869701385498,
+      "learning_rate": 1.1161161161161162e-06,
+      "loss": 0.3024,
+      "step": 88750
+    },
+    {
+      "epoch": 266.55,
+      "grad_norm": 20.32146453857422,
+      "learning_rate": 1.1151151151151152e-06,
+      "loss": 0.3173,
+      "step": 88760
+    },
+    {
+      "epoch": 266.58,
+      "grad_norm": 20.935014724731445,
+      "learning_rate": 1.1141141141141142e-06,
+      "loss": 0.3433,
+      "step": 88770
+    },
+    {
+      "epoch": 266.61,
+      "grad_norm": 21.355731964111328,
+      "learning_rate": 1.1131131131131132e-06,
+      "loss": 0.3106,
+      "step": 88780
+    },
+    {
+      "epoch": 266.64,
+      "grad_norm": 15.65914535522461,
+      "learning_rate": 1.1121121121121122e-06,
+      "loss": 0.31,
+      "step": 88790
+    },
+    {
+      "epoch": 266.67,
+      "grad_norm": 15.032240867614746,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.3328,
+      "step": 88800
+    },
+    {
+      "epoch": 266.7,
+      "grad_norm": 21.96764373779297,
+      "learning_rate": 1.11011011011011e-06,
+      "loss": 0.324,
+      "step": 88810
+    },
+    {
+      "epoch": 266.73,
+      "grad_norm": 14.819470405578613,
+      "learning_rate": 1.1091091091091093e-06,
+      "loss": 0.3392,
+      "step": 88820
+    },
+    {
+      "epoch": 266.76,
+      "grad_norm": 14.410465240478516,
+      "learning_rate": 1.1081081081081083e-06,
+      "loss": 0.3023,
+      "step": 88830
+    },
+    {
+      "epoch": 266.79,
+      "grad_norm": 11.120160102844238,
+      "learning_rate": 1.1071071071071073e-06,
+      "loss": 0.3367,
+      "step": 88840
+    },
+    {
+      "epoch": 266.82,
+      "grad_norm": 10.134851455688477,
+      "learning_rate": 1.1061061061061063e-06,
+      "loss": 0.2842,
+      "step": 88850
+    },
+    {
+      "epoch": 266.85,
+      "grad_norm": 19.21183967590332,
+      "learning_rate": 1.1051051051051051e-06,
+      "loss": 0.3263,
+      "step": 88860
+    },
+    {
+      "epoch": 266.88,
+      "grad_norm": 12.588943481445312,
+      "learning_rate": 1.1041041041041042e-06,
+      "loss": 0.3457,
+      "step": 88870
+    },
+    {
+      "epoch": 266.91,
+      "grad_norm": 13.658662796020508,
+      "learning_rate": 1.1031031031031032e-06,
+      "loss": 0.3233,
+      "step": 88880
+    },
+    {
+      "epoch": 266.94,
+      "grad_norm": 11.6569185256958,
+      "learning_rate": 1.1021021021021022e-06,
+      "loss": 0.3407,
+      "step": 88890
+    },
+    {
+      "epoch": 266.97,
+      "grad_norm": 13.841533660888672,
+      "learning_rate": 1.1011011011011012e-06,
+      "loss": 0.3426,
+      "step": 88900
+    },
+    {
+      "epoch": 267.0,
+      "grad_norm": 12.833606719970703,
+      "learning_rate": 1.1001001001001002e-06,
+      "loss": 0.3277,
+      "step": 88910
+    },
+    {
+      "epoch": 267.0,
+      "eval_accuracy": 0.8643,
+      "eval_loss": 0.5525545477867126,
+      "eval_runtime": 12.6325,
+      "eval_samples_per_second": 791.606,
+      "eval_steps_per_second": 3.166,
+      "step": 88911
+    },
+    {
+      "epoch": 267.03,
+      "grad_norm": 20.328781127929688,
+      "learning_rate": 1.0990990990990993e-06,
+      "loss": 0.3203,
+      "step": 88920
+    },
+    {
+      "epoch": 267.06,
+      "grad_norm": 13.420598983764648,
+      "learning_rate": 1.0980980980980983e-06,
+      "loss": 0.2995,
+      "step": 88930
+    },
+    {
+      "epoch": 267.09,
+      "grad_norm": 21.319461822509766,
+      "learning_rate": 1.0970970970970973e-06,
+      "loss": 0.3133,
+      "step": 88940
+    },
+    {
+      "epoch": 267.12,
+      "grad_norm": 12.325850486755371,
+      "learning_rate": 1.096096096096096e-06,
+      "loss": 0.3393,
+      "step": 88950
+    },
+    {
+      "epoch": 267.15,
+      "grad_norm": 15.582413673400879,
+      "learning_rate": 1.0950950950950951e-06,
+      "loss": 0.3137,
+      "step": 88960
+    },
+    {
+      "epoch": 267.18,
+      "grad_norm": 21.711448669433594,
+      "learning_rate": 1.0940940940940941e-06,
+      "loss": 0.3118,
+      "step": 88970
+    },
+    {
+      "epoch": 267.21,
+      "grad_norm": 20.662734985351562,
+      "learning_rate": 1.0930930930930932e-06,
+      "loss": 0.378,
+      "step": 88980
+    },
+    {
+      "epoch": 267.24,
+      "grad_norm": 15.549322128295898,
+      "learning_rate": 1.0920920920920922e-06,
+      "loss": 0.3431,
+      "step": 88990
+    },
+    {
+      "epoch": 267.27,
+      "grad_norm": 13.35905933380127,
+      "learning_rate": 1.0910910910910912e-06,
+      "loss": 0.3513,
+      "step": 89000
+    },
+    {
+      "epoch": 267.3,
+      "grad_norm": 10.23607063293457,
+      "learning_rate": 1.0900900900900902e-06,
+      "loss": 0.3428,
+      "step": 89010
+    },
+    {
+      "epoch": 267.33,
+      "grad_norm": 11.55119800567627,
+      "learning_rate": 1.0890890890890892e-06,
+      "loss": 0.3493,
+      "step": 89020
+    },
+    {
+      "epoch": 267.36,
+      "grad_norm": 12.0934419631958,
+      "learning_rate": 1.088088088088088e-06,
+      "loss": 0.3433,
+      "step": 89030
+    },
+    {
+      "epoch": 267.39,
+      "grad_norm": 14.261839866638184,
+      "learning_rate": 1.087087087087087e-06,
+      "loss": 0.2711,
+      "step": 89040
+    },
+    {
+      "epoch": 267.42,
+      "grad_norm": 14.679289817810059,
+      "learning_rate": 1.086086086086086e-06,
+      "loss": 0.3389,
+      "step": 89050
+    },
+    {
+      "epoch": 267.45,
+      "grad_norm": 18.5410213470459,
+      "learning_rate": 1.085085085085085e-06,
+      "loss": 0.3044,
+      "step": 89060
+    },
+    {
+      "epoch": 267.48,
+      "grad_norm": 12.537636756896973,
+      "learning_rate": 1.0840840840840841e-06,
+      "loss": 0.2941,
+      "step": 89070
+    },
+    {
+      "epoch": 267.51,
+      "grad_norm": 14.628470420837402,
+      "learning_rate": 1.0830830830830831e-06,
+      "loss": 0.3752,
+      "step": 89080
+    },
+    {
+      "epoch": 267.54,
+      "grad_norm": 13.823002815246582,
+      "learning_rate": 1.0820820820820821e-06,
+      "loss": 0.3396,
+      "step": 89090
+    },
+    {
+      "epoch": 267.57,
+      "grad_norm": 14.301614761352539,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.3659,
+      "step": 89100
+    },
+    {
+      "epoch": 267.6,
+      "grad_norm": 13.403546333312988,
+      "learning_rate": 1.0800800800800802e-06,
+      "loss": 0.3574,
+      "step": 89110
+    },
+    {
+      "epoch": 267.63,
+      "grad_norm": 15.12373161315918,
+      "learning_rate": 1.0790790790790792e-06,
+      "loss": 0.3065,
+      "step": 89120
+    },
+    {
+      "epoch": 267.66,
+      "grad_norm": 16.834741592407227,
+      "learning_rate": 1.0780780780780782e-06,
+      "loss": 0.3857,
+      "step": 89130
+    },
+    {
+      "epoch": 267.69,
+      "grad_norm": 13.995725631713867,
+      "learning_rate": 1.0770770770770772e-06,
+      "loss": 0.2978,
+      "step": 89140
+    },
+    {
+      "epoch": 267.72,
+      "grad_norm": 11.281320571899414,
+      "learning_rate": 1.0760760760760763e-06,
+      "loss": 0.3007,
+      "step": 89150
+    },
+    {
+      "epoch": 267.75,
+      "grad_norm": 14.01447868347168,
+      "learning_rate": 1.0750750750750753e-06,
+      "loss": 0.3136,
+      "step": 89160
+    },
+    {
+      "epoch": 267.78,
+      "grad_norm": 21.85301971435547,
+      "learning_rate": 1.074074074074074e-06,
+      "loss": 0.357,
+      "step": 89170
+    },
+    {
+      "epoch": 267.81,
+      "grad_norm": 9.722794532775879,
+      "learning_rate": 1.073073073073073e-06,
+      "loss": 0.2999,
+      "step": 89180
+    },
+    {
+      "epoch": 267.84,
+      "grad_norm": 22.79783821105957,
+      "learning_rate": 1.0720720720720721e-06,
+      "loss": 0.3307,
+      "step": 89190
+    },
+    {
+      "epoch": 267.87,
+      "grad_norm": 21.149213790893555,
+      "learning_rate": 1.0710710710710711e-06,
+      "loss": 0.3184,
+      "step": 89200
+    },
+    {
+      "epoch": 267.9,
+      "grad_norm": 12.08806037902832,
+      "learning_rate": 1.0700700700700702e-06,
+      "loss": 0.33,
+      "step": 89210
+    },
+    {
+      "epoch": 267.93,
+      "grad_norm": 15.27159309387207,
+      "learning_rate": 1.0690690690690692e-06,
+      "loss": 0.3234,
+      "step": 89220
+    },
+    {
+      "epoch": 267.96,
+      "grad_norm": 19.168285369873047,
+      "learning_rate": 1.0680680680680682e-06,
+      "loss": 0.3575,
+      "step": 89230
+    },
+    {
+      "epoch": 267.99,
+      "grad_norm": 15.897375106811523,
+      "learning_rate": 1.0670670670670672e-06,
+      "loss": 0.3304,
+      "step": 89240
+    },
+    {
+      "epoch": 268.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.548311710357666,
+      "eval_runtime": 12.9065,
+      "eval_samples_per_second": 774.803,
+      "eval_steps_per_second": 3.099,
+      "step": 89244
+    },
+    {
+      "epoch": 268.02,
+      "grad_norm": 22.204694747924805,
+      "learning_rate": 1.0660660660660662e-06,
+      "loss": 0.3378,
+      "step": 89250
+    },
+    {
+      "epoch": 268.05,
+      "grad_norm": 18.930078506469727,
+      "learning_rate": 1.065065065065065e-06,
+      "loss": 0.3283,
+      "step": 89260
+    },
+    {
+      "epoch": 268.08,
+      "grad_norm": 11.978702545166016,
+      "learning_rate": 1.064064064064064e-06,
+      "loss": 0.292,
+      "step": 89270
+    },
+    {
+      "epoch": 268.11,
+      "grad_norm": 12.391798973083496,
+      "learning_rate": 1.063063063063063e-06,
+      "loss": 0.3294,
+      "step": 89280
+    },
+    {
+      "epoch": 268.14,
+      "grad_norm": 9.46595573425293,
+      "learning_rate": 1.062062062062062e-06,
+      "loss": 0.3175,
+      "step": 89290
+    },
+    {
+      "epoch": 268.17,
+      "grad_norm": 18.108123779296875,
+      "learning_rate": 1.0610610610610611e-06,
+      "loss": 0.3706,
+      "step": 89300
+    },
+    {
+      "epoch": 268.2,
+      "grad_norm": 11.40620231628418,
+      "learning_rate": 1.0600600600600601e-06,
+      "loss": 0.3397,
+      "step": 89310
+    },
+    {
+      "epoch": 268.23,
+      "grad_norm": 22.594858169555664,
+      "learning_rate": 1.0590590590590591e-06,
+      "loss": 0.3419,
+      "step": 89320
+    },
+    {
+      "epoch": 268.26,
+      "grad_norm": 22.711406707763672,
+      "learning_rate": 1.0580580580580582e-06,
+      "loss": 0.3251,
+      "step": 89330
+    },
+    {
+      "epoch": 268.29,
+      "grad_norm": 16.294572830200195,
+      "learning_rate": 1.0570570570570572e-06,
+      "loss": 0.3487,
+      "step": 89340
+    },
+    {
+      "epoch": 268.32,
+      "grad_norm": 14.993041038513184,
+      "learning_rate": 1.0560560560560562e-06,
+      "loss": 0.3347,
+      "step": 89350
+    },
+    {
+      "epoch": 268.35,
+      "grad_norm": 13.742610931396484,
+      "learning_rate": 1.0550550550550552e-06,
+      "loss": 0.3733,
+      "step": 89360
+    },
+    {
+      "epoch": 268.38,
+      "grad_norm": 9.686351776123047,
+      "learning_rate": 1.0540540540540542e-06,
+      "loss": 0.3203,
+      "step": 89370
+    },
+    {
+      "epoch": 268.41,
+      "grad_norm": 14.923864364624023,
+      "learning_rate": 1.0530530530530533e-06,
+      "loss": 0.3086,
+      "step": 89380
+    },
+    {
+      "epoch": 268.44,
+      "grad_norm": 12.566535949707031,
+      "learning_rate": 1.0520520520520523e-06,
+      "loss": 0.349,
+      "step": 89390
+    },
+    {
+      "epoch": 268.47,
+      "grad_norm": 16.487207412719727,
+      "learning_rate": 1.051051051051051e-06,
+      "loss": 0.2804,
+      "step": 89400
+    },
+    {
+      "epoch": 268.5,
+      "grad_norm": 27.251853942871094,
+      "learning_rate": 1.05005005005005e-06,
+      "loss": 0.3061,
+      "step": 89410
+    },
+    {
+      "epoch": 268.53,
+      "grad_norm": 15.900343894958496,
+      "learning_rate": 1.0490490490490491e-06,
+      "loss": 0.3565,
+      "step": 89420
+    },
+    {
+      "epoch": 268.56,
+      "grad_norm": 10.381702423095703,
+      "learning_rate": 1.0480480480480481e-06,
+      "loss": 0.3392,
+      "step": 89430
+    },
+    {
+      "epoch": 268.59,
+      "grad_norm": 16.311491012573242,
+      "learning_rate": 1.0470470470470472e-06,
+      "loss": 0.322,
+      "step": 89440
+    },
+    {
+      "epoch": 268.62,
+      "grad_norm": 19.335205078125,
+      "learning_rate": 1.0460460460460462e-06,
+      "loss": 0.3608,
+      "step": 89450
+    },
+    {
+      "epoch": 268.65,
+      "grad_norm": 27.31247329711914,
+      "learning_rate": 1.0450450450450452e-06,
+      "loss": 0.3669,
+      "step": 89460
+    },
+    {
+      "epoch": 268.68,
+      "grad_norm": 16.436534881591797,
+      "learning_rate": 1.0440440440440442e-06,
+      "loss": 0.3444,
+      "step": 89470
+    },
+    {
+      "epoch": 268.71,
+      "grad_norm": 17.860103607177734,
+      "learning_rate": 1.0430430430430432e-06,
+      "loss": 0.3447,
+      "step": 89480
+    },
+    {
+      "epoch": 268.74,
+      "grad_norm": 18.588132858276367,
+      "learning_rate": 1.042042042042042e-06,
+      "loss": 0.2799,
+      "step": 89490
+    },
+    {
+      "epoch": 268.77,
+      "grad_norm": 13.699021339416504,
+      "learning_rate": 1.041041041041041e-06,
+      "loss": 0.326,
+      "step": 89500
+    },
+    {
+      "epoch": 268.8,
+      "grad_norm": 17.24247932434082,
+      "learning_rate": 1.04004004004004e-06,
+      "loss": 0.3212,
+      "step": 89510
+    },
+    {
+      "epoch": 268.83,
+      "grad_norm": 11.796412467956543,
+      "learning_rate": 1.039039039039039e-06,
+      "loss": 0.3272,
+      "step": 89520
+    },
+    {
+      "epoch": 268.86,
+      "grad_norm": 17.92228126525879,
+      "learning_rate": 1.038038038038038e-06,
+      "loss": 0.3347,
+      "step": 89530
+    },
+    {
+      "epoch": 268.89,
+      "grad_norm": 18.643171310424805,
+      "learning_rate": 1.0370370370370371e-06,
+      "loss": 0.2527,
+      "step": 89540
+    },
+    {
+      "epoch": 268.92,
+      "grad_norm": 11.988188743591309,
+      "learning_rate": 1.0360360360360361e-06,
+      "loss": 0.3055,
+      "step": 89550
+    },
+    {
+      "epoch": 268.95,
+      "grad_norm": 18.95526885986328,
+      "learning_rate": 1.0350350350350352e-06,
+      "loss": 0.3518,
+      "step": 89560
+    },
+    {
+      "epoch": 268.98,
+      "grad_norm": 11.828939437866211,
+      "learning_rate": 1.0340340340340342e-06,
+      "loss": 0.2967,
+      "step": 89570
+    },
+    {
+      "epoch": 269.0,
+      "eval_accuracy": 0.8642,
+      "eval_loss": 0.548523485660553,
+      "eval_runtime": 12.8268,
+      "eval_samples_per_second": 779.621,
+      "eval_steps_per_second": 3.118,
+      "step": 89577
+    },
+    {
+      "epoch": 269.01,
+      "grad_norm": 14.977457046508789,
+      "learning_rate": 1.033033033033033e-06,
+      "loss": 0.3069,
+      "step": 89580
+    },
+    {
+      "epoch": 269.04,
+      "grad_norm": 13.671182632446289,
+      "learning_rate": 1.032032032032032e-06,
+      "loss": 0.3227,
+      "step": 89590
+    },
+    {
+      "epoch": 269.07,
+      "grad_norm": 16.29219627380371,
+      "learning_rate": 1.031031031031031e-06,
+      "loss": 0.3451,
+      "step": 89600
+    },
+    {
+      "epoch": 269.1,
+      "grad_norm": 10.795699119567871,
+      "learning_rate": 1.03003003003003e-06,
+      "loss": 0.3397,
+      "step": 89610
+    },
+    {
+      "epoch": 269.13,
+      "grad_norm": 18.84855842590332,
+      "learning_rate": 1.0290290290290293e-06,
+      "loss": 0.3467,
+      "step": 89620
+    },
+    {
+      "epoch": 269.16,
+      "grad_norm": 13.698019981384277,
+      "learning_rate": 1.028028028028028e-06,
+      "loss": 0.3285,
+      "step": 89630
+    },
+    {
+      "epoch": 269.19,
+      "grad_norm": 18.709293365478516,
+      "learning_rate": 1.027027027027027e-06,
+      "loss": 0.325,
+      "step": 89640
+    },
+    {
+      "epoch": 269.22,
+      "grad_norm": 12.8465576171875,
+      "learning_rate": 1.0260260260260261e-06,
+      "loss": 0.3769,
+      "step": 89650
+    },
+    {
+      "epoch": 269.25,
+      "grad_norm": 17.539836883544922,
+      "learning_rate": 1.0250250250250251e-06,
+      "loss": 0.3178,
+      "step": 89660
+    },
+    {
+      "epoch": 269.28,
+      "grad_norm": 16.641210556030273,
+      "learning_rate": 1.0240240240240242e-06,
+      "loss": 0.3333,
+      "step": 89670
+    },
+    {
+      "epoch": 269.31,
+      "grad_norm": 22.230295181274414,
+      "learning_rate": 1.0230230230230232e-06,
+      "loss": 0.35,
+      "step": 89680
+    },
+    {
+      "epoch": 269.34,
+      "grad_norm": 16.390005111694336,
+      "learning_rate": 1.0220220220220222e-06,
+      "loss": 0.2899,
+      "step": 89690
+    },
+    {
+      "epoch": 269.37,
+      "grad_norm": 20.934371948242188,
+      "learning_rate": 1.0210210210210212e-06,
+      "loss": 0.3218,
+      "step": 89700
+    },
+    {
+      "epoch": 269.4,
+      "grad_norm": 12.79007625579834,
+      "learning_rate": 1.0200200200200202e-06,
+      "loss": 0.2614,
+      "step": 89710
+    },
+    {
+      "epoch": 269.43,
+      "grad_norm": 21.067394256591797,
+      "learning_rate": 1.019019019019019e-06,
+      "loss": 0.3622,
+      "step": 89720
+    },
+    {
+      "epoch": 269.46,
+      "grad_norm": 15.322122573852539,
+      "learning_rate": 1.018018018018018e-06,
+      "loss": 0.3375,
+      "step": 89730
+    },
+    {
+      "epoch": 269.49,
+      "grad_norm": 14.134215354919434,
+      "learning_rate": 1.017017017017017e-06,
+      "loss": 0.3085,
+      "step": 89740
+    },
+    {
+      "epoch": 269.52,
+      "grad_norm": 21.32209014892578,
+      "learning_rate": 1.016016016016016e-06,
+      "loss": 0.2971,
+      "step": 89750
+    },
+    {
+      "epoch": 269.55,
+      "grad_norm": 14.173908233642578,
+      "learning_rate": 1.015015015015015e-06,
+      "loss": 0.3556,
+      "step": 89760
+    },
+    {
+      "epoch": 269.58,
+      "grad_norm": 20.633502960205078,
+      "learning_rate": 1.0140140140140141e-06,
+      "loss": 0.2915,
+      "step": 89770
+    },
+    {
+      "epoch": 269.61,
+      "grad_norm": 13.786351203918457,
+      "learning_rate": 1.0130130130130131e-06,
+      "loss": 0.3046,
+      "step": 89780
+    },
+    {
+      "epoch": 269.64,
+      "grad_norm": 15.203786849975586,
+      "learning_rate": 1.0120120120120122e-06,
+      "loss": 0.3151,
+      "step": 89790
+    },
+    {
+      "epoch": 269.67,
+      "grad_norm": 14.01477336883545,
+      "learning_rate": 1.011011011011011e-06,
+      "loss": 0.3449,
+      "step": 89800
+    },
+    {
+      "epoch": 269.7,
+      "grad_norm": 22.405813217163086,
+      "learning_rate": 1.01001001001001e-06,
+      "loss": 0.3515,
+      "step": 89810
+    },
+    {
+      "epoch": 269.73,
+      "grad_norm": 21.73993492126465,
+      "learning_rate": 1.009009009009009e-06,
+      "loss": 0.3256,
+      "step": 89820
+    },
+    {
+      "epoch": 269.76,
+      "grad_norm": 28.394853591918945,
+      "learning_rate": 1.008008008008008e-06,
+      "loss": 0.2816,
+      "step": 89830
+    },
+    {
+      "epoch": 269.79,
+      "grad_norm": 21.098861694335938,
+      "learning_rate": 1.007007007007007e-06,
+      "loss": 0.3629,
+      "step": 89840
+    },
+    {
+      "epoch": 269.82,
+      "grad_norm": 12.8715238571167,
+      "learning_rate": 1.006006006006006e-06,
+      "loss": 0.295,
+      "step": 89850
+    },
+    {
+      "epoch": 269.85,
+      "grad_norm": 12.514232635498047,
+      "learning_rate": 1.005005005005005e-06,
+      "loss": 0.3432,
+      "step": 89860
+    },
+    {
+      "epoch": 269.88,
+      "grad_norm": 18.550094604492188,
+      "learning_rate": 1.004004004004004e-06,
+      "loss": 0.3559,
+      "step": 89870
+    },
+    {
+      "epoch": 269.91,
+      "grad_norm": 14.617057800292969,
+      "learning_rate": 1.0030030030030031e-06,
+      "loss": 0.362,
+      "step": 89880
+    },
+    {
+      "epoch": 269.94,
+      "grad_norm": 21.027708053588867,
+      "learning_rate": 1.0020020020020021e-06,
+      "loss": 0.3837,
+      "step": 89890
+    },
+    {
+      "epoch": 269.97,
+      "grad_norm": 24.867334365844727,
+      "learning_rate": 1.0010010010010011e-06,
+      "loss": 0.3715,
+      "step": 89900
+    },
+    {
+      "epoch": 270.0,
+      "grad_norm": 17.234952926635742,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.2956,
+      "step": 89910
+    },
+    {
+      "epoch": 270.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5505730509757996,
+      "eval_runtime": 13.0192,
+      "eval_samples_per_second": 768.095,
+      "eval_steps_per_second": 3.072,
+      "step": 89910
+    },
+    {
+      "epoch": 270.03,
+      "grad_norm": 13.731232643127441,
+      "learning_rate": 9.989989989989992e-07,
+      "loss": 0.3446,
+      "step": 89920
+    },
+    {
+      "epoch": 270.06,
+      "grad_norm": 13.255059242248535,
+      "learning_rate": 9.979979979979982e-07,
+      "loss": 0.2957,
+      "step": 89930
+    },
+    {
+      "epoch": 270.09,
+      "grad_norm": 16.162519454956055,
+      "learning_rate": 9.969969969969972e-07,
+      "loss": 0.3512,
+      "step": 89940
+    },
+    {
+      "epoch": 270.12,
+      "grad_norm": 16.619470596313477,
+      "learning_rate": 9.95995995995996e-07,
+      "loss": 0.3318,
+      "step": 89950
+    },
+    {
+      "epoch": 270.15,
+      "grad_norm": 13.50727367401123,
+      "learning_rate": 9.94994994994995e-07,
+      "loss": 0.3323,
+      "step": 89960
+    },
+    {
+      "epoch": 270.18,
+      "grad_norm": 19.635948181152344,
+      "learning_rate": 9.93993993993994e-07,
+      "loss": 0.3134,
+      "step": 89970
+    },
+    {
+      "epoch": 270.21,
+      "grad_norm": 20.472583770751953,
+      "learning_rate": 9.92992992992993e-07,
+      "loss": 0.3665,
+      "step": 89980
+    },
+    {
+      "epoch": 270.24,
+      "grad_norm": 20.509553909301758,
+      "learning_rate": 9.91991991991992e-07,
+      "loss": 0.2937,
+      "step": 89990
+    },
+    {
+      "epoch": 270.27,
+      "grad_norm": 11.262863159179688,
+      "learning_rate": 9.909909909909911e-07,
+      "loss": 0.3283,
+      "step": 90000
+    },
+    {
+      "epoch": 270.3,
+      "grad_norm": 13.4719877243042,
+      "learning_rate": 9.899899899899901e-07,
+      "loss": 0.2946,
+      "step": 90010
+    },
+    {
+      "epoch": 270.33,
+      "grad_norm": 22.185752868652344,
+      "learning_rate": 9.889889889889892e-07,
+      "loss": 0.323,
+      "step": 90020
+    },
+    {
+      "epoch": 270.36,
+      "grad_norm": 21.662609100341797,
+      "learning_rate": 9.87987987987988e-07,
+      "loss": 0.3194,
+      "step": 90030
+    },
+    {
+      "epoch": 270.39,
+      "grad_norm": 16.972679138183594,
+      "learning_rate": 9.86986986986987e-07,
+      "loss": 0.3077,
+      "step": 90040
+    },
+    {
+      "epoch": 270.42,
+      "grad_norm": 15.681986808776855,
+      "learning_rate": 9.85985985985986e-07,
+      "loss": 0.3213,
+      "step": 90050
+    },
+    {
+      "epoch": 270.45,
+      "grad_norm": 15.448613166809082,
+      "learning_rate": 9.84984984984985e-07,
+      "loss": 0.3353,
+      "step": 90060
+    },
+    {
+      "epoch": 270.48,
+      "grad_norm": 11.712885856628418,
+      "learning_rate": 9.83983983983984e-07,
+      "loss": 0.3292,
+      "step": 90070
+    },
+    {
+      "epoch": 270.51,
+      "grad_norm": 14.702363014221191,
+      "learning_rate": 9.82982982982983e-07,
+      "loss": 0.3532,
+      "step": 90080
+    },
+    {
+      "epoch": 270.54,
+      "grad_norm": 18.324710845947266,
+      "learning_rate": 9.81981981981982e-07,
+      "loss": 0.3254,
+      "step": 90090
+    },
+    {
+      "epoch": 270.57,
+      "grad_norm": 18.324268341064453,
+      "learning_rate": 9.80980980980981e-07,
+      "loss": 0.3131,
+      "step": 90100
+    },
+    {
+      "epoch": 270.6,
+      "grad_norm": 25.963987350463867,
+      "learning_rate": 9.799799799799801e-07,
+      "loss": 0.3567,
+      "step": 90110
+    },
+    {
+      "epoch": 270.63,
+      "grad_norm": 18.823331832885742,
+      "learning_rate": 9.78978978978979e-07,
+      "loss": 0.3559,
+      "step": 90120
+    },
+    {
+      "epoch": 270.66,
+      "grad_norm": 17.104080200195312,
+      "learning_rate": 9.77977977977978e-07,
+      "loss": 0.3838,
+      "step": 90130
+    },
+    {
+      "epoch": 270.69,
+      "grad_norm": 11.049248695373535,
+      "learning_rate": 9.76976976976977e-07,
+      "loss": 0.3066,
+      "step": 90140
+    },
+    {
+      "epoch": 270.72,
+      "grad_norm": 10.526016235351562,
+      "learning_rate": 9.75975975975976e-07,
+      "loss": 0.325,
+      "step": 90150
+    },
+    {
+      "epoch": 270.75,
+      "grad_norm": 21.767608642578125,
+      "learning_rate": 9.749749749749752e-07,
+      "loss": 0.3584,
+      "step": 90160
+    },
+    {
+      "epoch": 270.78,
+      "grad_norm": 18.4510555267334,
+      "learning_rate": 9.73973973973974e-07,
+      "loss": 0.3103,
+      "step": 90170
+    },
+    {
+      "epoch": 270.81,
+      "grad_norm": 17.46860122680664,
+      "learning_rate": 9.72972972972973e-07,
+      "loss": 0.3074,
+      "step": 90180
+    },
+    {
+      "epoch": 270.84,
+      "grad_norm": 15.136173248291016,
+      "learning_rate": 9.71971971971972e-07,
+      "loss": 0.2953,
+      "step": 90190
+    },
+    {
+      "epoch": 270.87,
+      "grad_norm": 13.894150733947754,
+      "learning_rate": 9.70970970970971e-07,
+      "loss": 0.3475,
+      "step": 90200
+    },
+    {
+      "epoch": 270.9,
+      "grad_norm": 11.757837295532227,
+      "learning_rate": 9.6996996996997e-07,
+      "loss": 0.3362,
+      "step": 90210
+    },
+    {
+      "epoch": 270.93,
+      "grad_norm": 27.379348754882812,
+      "learning_rate": 9.68968968968969e-07,
+      "loss": 0.3977,
+      "step": 90220
+    },
+    {
+      "epoch": 270.96,
+      "grad_norm": 12.738484382629395,
+      "learning_rate": 9.679679679679681e-07,
+      "loss": 0.3286,
+      "step": 90230
+    },
+    {
+      "epoch": 270.99,
+      "grad_norm": 17.733993530273438,
+      "learning_rate": 9.669669669669671e-07,
+      "loss": 0.335,
+      "step": 90240
+    },
+    {
+      "epoch": 271.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.5496519804000854,
+      "eval_runtime": 12.4188,
+      "eval_samples_per_second": 805.233,
+      "eval_steps_per_second": 3.221,
+      "step": 90243
+    },
+    {
+      "epoch": 271.02,
+      "grad_norm": 18.647645950317383,
+      "learning_rate": 9.659659659659662e-07,
+      "loss": 0.3313,
+      "step": 90250
+    },
+    {
+      "epoch": 271.05,
+      "grad_norm": 16.820722579956055,
+      "learning_rate": 9.64964964964965e-07,
+      "loss": 0.2973,
+      "step": 90260
+    },
+    {
+      "epoch": 271.08,
+      "grad_norm": 15.603100776672363,
+      "learning_rate": 9.63963963963964e-07,
+      "loss": 0.3289,
+      "step": 90270
+    },
+    {
+      "epoch": 271.11,
+      "grad_norm": 12.769047737121582,
+      "learning_rate": 9.62962962962963e-07,
+      "loss": 0.3102,
+      "step": 90280
+    },
+    {
+      "epoch": 271.14,
+      "grad_norm": 16.098630905151367,
+      "learning_rate": 9.61961961961962e-07,
+      "loss": 0.3128,
+      "step": 90290
+    },
+    {
+      "epoch": 271.17,
+      "grad_norm": 15.355938911437988,
+      "learning_rate": 9.60960960960961e-07,
+      "loss": 0.3154,
+      "step": 90300
+    },
+    {
+      "epoch": 271.2,
+      "grad_norm": 29.035205841064453,
+      "learning_rate": 9.5995995995996e-07,
+      "loss": 0.3312,
+      "step": 90310
+    },
+    {
+      "epoch": 271.23,
+      "grad_norm": 17.579605102539062,
+      "learning_rate": 9.58958958958959e-07,
+      "loss": 0.3344,
+      "step": 90320
+    },
+    {
+      "epoch": 271.26,
+      "grad_norm": 11.887344360351562,
+      "learning_rate": 9.57957957957958e-07,
+      "loss": 0.3561,
+      "step": 90330
+    },
+    {
+      "epoch": 271.29,
+      "grad_norm": 20.4341983795166,
+      "learning_rate": 9.569569569569571e-07,
+      "loss": 0.3809,
+      "step": 90340
+    },
+    {
+      "epoch": 271.32,
+      "grad_norm": 16.298974990844727,
+      "learning_rate": 9.55955955955956e-07,
+      "loss": 0.3295,
+      "step": 90350
+    },
+    {
+      "epoch": 271.35,
+      "grad_norm": 19.652822494506836,
+      "learning_rate": 9.54954954954955e-07,
+      "loss": 0.3305,
+      "step": 90360
+    },
+    {
+      "epoch": 271.38,
+      "grad_norm": 13.072126388549805,
+      "learning_rate": 9.53953953953954e-07,
+      "loss": 0.3058,
+      "step": 90370
+    },
+    {
+      "epoch": 271.41,
+      "grad_norm": 9.867013931274414,
+      "learning_rate": 9.52952952952953e-07,
+      "loss": 0.3488,
+      "step": 90380
+    },
+    {
+      "epoch": 271.44,
+      "grad_norm": 15.455288887023926,
+      "learning_rate": 9.51951951951952e-07,
+      "loss": 0.3403,
+      "step": 90390
+    },
+    {
+      "epoch": 271.47,
+      "grad_norm": 18.56848907470703,
+      "learning_rate": 9.50950950950951e-07,
+      "loss": 0.3134,
+      "step": 90400
+    },
+    {
+      "epoch": 271.5,
+      "grad_norm": 10.916102409362793,
+      "learning_rate": 9.499499499499499e-07,
+      "loss": 0.3051,
+      "step": 90410
+    },
+    {
+      "epoch": 271.53,
+      "grad_norm": 12.692329406738281,
+      "learning_rate": 9.48948948948949e-07,
+      "loss": 0.2754,
+      "step": 90420
+    },
+    {
+      "epoch": 271.56,
+      "grad_norm": 17.429264068603516,
+      "learning_rate": 9.479479479479481e-07,
+      "loss": 0.3375,
+      "step": 90430
+    },
+    {
+      "epoch": 271.59,
+      "grad_norm": 15.141221046447754,
+      "learning_rate": 9.469469469469471e-07,
+      "loss": 0.2709,
+      "step": 90440
+    },
+    {
+      "epoch": 271.62,
+      "grad_norm": 13.29561710357666,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 0.3282,
+      "step": 90450
+    },
+    {
+      "epoch": 271.65,
+      "grad_norm": 13.768024444580078,
+      "learning_rate": 9.449449449449451e-07,
+      "loss": 0.296,
+      "step": 90460
+    },
+    {
+      "epoch": 271.68,
+      "grad_norm": 14.572839736938477,
+      "learning_rate": 9.43943943943944e-07,
+      "loss": 0.3388,
+      "step": 90470
+    },
+    {
+      "epoch": 271.71,
+      "grad_norm": 14.353155136108398,
+      "learning_rate": 9.42942942942943e-07,
+      "loss": 0.2977,
+      "step": 90480
+    },
+    {
+      "epoch": 271.74,
+      "grad_norm": 15.48045539855957,
+      "learning_rate": 9.419419419419421e-07,
+      "loss": 0.3694,
+      "step": 90490
+    },
+    {
+      "epoch": 271.77,
+      "grad_norm": 12.211432456970215,
+      "learning_rate": 9.409409409409411e-07,
+      "loss": 0.2845,
+      "step": 90500
+    },
+    {
+      "epoch": 271.8,
+      "grad_norm": 10.956110000610352,
+      "learning_rate": 9.3993993993994e-07,
+      "loss": 0.3093,
+      "step": 90510
+    },
+    {
+      "epoch": 271.83,
+      "grad_norm": 10.659209251403809,
+      "learning_rate": 9.38938938938939e-07,
+      "loss": 0.3262,
+      "step": 90520
+    },
+    {
+      "epoch": 271.86,
+      "grad_norm": 10.500244140625,
+      "learning_rate": 9.37937937937938e-07,
+      "loss": 0.2767,
+      "step": 90530
+    },
+    {
+      "epoch": 271.89,
+      "grad_norm": 15.689346313476562,
+      "learning_rate": 9.36936936936937e-07,
+      "loss": 0.3347,
+      "step": 90540
+    },
+    {
+      "epoch": 271.92,
+      "grad_norm": 13.565827369689941,
+      "learning_rate": 9.35935935935936e-07,
+      "loss": 0.269,
+      "step": 90550
+    },
+    {
+      "epoch": 271.95,
+      "grad_norm": 15.301671028137207,
+      "learning_rate": 9.34934934934935e-07,
+      "loss": 0.3265,
+      "step": 90560
+    },
+    {
+      "epoch": 271.98,
+      "grad_norm": 13.416406631469727,
+      "learning_rate": 9.33933933933934e-07,
+      "loss": 0.3032,
+      "step": 90570
+    },
+    {
+      "epoch": 272.0,
+      "eval_accuracy": 0.8642,
+      "eval_loss": 0.549430787563324,
+      "eval_runtime": 12.7522,
+      "eval_samples_per_second": 784.176,
+      "eval_steps_per_second": 3.137,
+      "step": 90576
+    },
+    {
+      "epoch": 272.01,
+      "grad_norm": 9.975958824157715,
+      "learning_rate": 9.32932932932933e-07,
+      "loss": 0.2643,
+      "step": 90580
+    },
+    {
+      "epoch": 272.04,
+      "grad_norm": 17.208833694458008,
+      "learning_rate": 9.31931931931932e-07,
+      "loss": 0.3287,
+      "step": 90590
+    },
+    {
+      "epoch": 272.07,
+      "grad_norm": 21.038414001464844,
+      "learning_rate": 9.30930930930931e-07,
+      "loss": 0.3485,
+      "step": 90600
+    },
+    {
+      "epoch": 272.1,
+      "grad_norm": 20.560436248779297,
+      "learning_rate": 9.2992992992993e-07,
+      "loss": 0.346,
+      "step": 90610
+    },
+    {
+      "epoch": 272.13,
+      "grad_norm": 11.736292839050293,
+      "learning_rate": 9.28928928928929e-07,
+      "loss": 0.3711,
+      "step": 90620
+    },
+    {
+      "epoch": 272.16,
+      "grad_norm": 13.6312255859375,
+      "learning_rate": 9.27927927927928e-07,
+      "loss": 0.2937,
+      "step": 90630
+    },
+    {
+      "epoch": 272.19,
+      "grad_norm": 13.546832084655762,
+      "learning_rate": 9.269269269269269e-07,
+      "loss": 0.3023,
+      "step": 90640
+    },
+    {
+      "epoch": 272.22,
+      "grad_norm": 15.164703369140625,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 0.3099,
+      "step": 90650
+    },
+    {
+      "epoch": 272.25,
+      "grad_norm": 18.45989227294922,
+      "learning_rate": 9.24924924924925e-07,
+      "loss": 0.3138,
+      "step": 90660
+    },
+    {
+      "epoch": 272.28,
+      "grad_norm": 20.216459274291992,
+      "learning_rate": 9.23923923923924e-07,
+      "loss": 0.305,
+      "step": 90670
+    },
+    {
+      "epoch": 272.31,
+      "grad_norm": 10.366315841674805,
+      "learning_rate": 9.229229229229229e-07,
+      "loss": 0.3275,
+      "step": 90680
+    },
+    {
+      "epoch": 272.34,
+      "grad_norm": 17.649259567260742,
+      "learning_rate": 9.219219219219221e-07,
+      "loss": 0.3452,
+      "step": 90690
+    },
+    {
+      "epoch": 272.37,
+      "grad_norm": 24.05565643310547,
+      "learning_rate": 9.20920920920921e-07,
+      "loss": 0.3197,
+      "step": 90700
+    },
+    {
+      "epoch": 272.4,
+      "grad_norm": 27.948877334594727,
+      "learning_rate": 9.1991991991992e-07,
+      "loss": 0.3345,
+      "step": 90710
+    },
+    {
+      "epoch": 272.43,
+      "grad_norm": 55.0609245300293,
+      "learning_rate": 9.189189189189191e-07,
+      "loss": 0.3406,
+      "step": 90720
+    },
+    {
+      "epoch": 272.46,
+      "grad_norm": 24.96762466430664,
+      "learning_rate": 9.179179179179181e-07,
+      "loss": 0.3089,
+      "step": 90730
+    },
+    {
+      "epoch": 272.49,
+      "grad_norm": 13.768011093139648,
+      "learning_rate": 9.16916916916917e-07,
+      "loss": 0.3556,
+      "step": 90740
+    },
+    {
+      "epoch": 272.52,
+      "grad_norm": 13.462057113647461,
+      "learning_rate": 9.15915915915916e-07,
+      "loss": 0.315,
+      "step": 90750
+    },
+    {
+      "epoch": 272.55,
+      "grad_norm": 21.621870040893555,
+      "learning_rate": 9.14914914914915e-07,
+      "loss": 0.3349,
+      "step": 90760
+    },
+    {
+      "epoch": 272.58,
+      "grad_norm": 13.644549369812012,
+      "learning_rate": 9.13913913913914e-07,
+      "loss": 0.3547,
+      "step": 90770
+    },
+    {
+      "epoch": 272.61,
+      "grad_norm": 20.444927215576172,
+      "learning_rate": 9.12912912912913e-07,
+      "loss": 0.3359,
+      "step": 90780
+    },
+    {
+      "epoch": 272.64,
+      "grad_norm": 14.784795761108398,
+      "learning_rate": 9.11911911911912e-07,
+      "loss": 0.4208,
+      "step": 90790
+    },
+    {
+      "epoch": 272.67,
+      "grad_norm": 17.933008193969727,
+      "learning_rate": 9.10910910910911e-07,
+      "loss": 0.3079,
+      "step": 90800
+    },
+    {
+      "epoch": 272.7,
+      "grad_norm": 10.328374862670898,
+      "learning_rate": 9.0990990990991e-07,
+      "loss": 0.2981,
+      "step": 90810
+    },
+    {
+      "epoch": 272.73,
+      "grad_norm": 16.268577575683594,
+      "learning_rate": 9.08908908908909e-07,
+      "loss": 0.3475,
+      "step": 90820
+    },
+    {
+      "epoch": 272.76,
+      "grad_norm": 14.276102066040039,
+      "learning_rate": 9.07907907907908e-07,
+      "loss": 0.3571,
+      "step": 90830
+    },
+    {
+      "epoch": 272.79,
+      "grad_norm": 21.006818771362305,
+      "learning_rate": 9.06906906906907e-07,
+      "loss": 0.367,
+      "step": 90840
+    },
+    {
+      "epoch": 272.82,
+      "grad_norm": 17.40716552734375,
+      "learning_rate": 9.05905905905906e-07,
+      "loss": 0.349,
+      "step": 90850
+    },
+    {
+      "epoch": 272.85,
+      "grad_norm": 11.275979995727539,
+      "learning_rate": 9.04904904904905e-07,
+      "loss": 0.2996,
+      "step": 90860
+    },
+    {
+      "epoch": 272.88,
+      "grad_norm": 10.080249786376953,
+      "learning_rate": 9.039039039039039e-07,
+      "loss": 0.3245,
+      "step": 90870
+    },
+    {
+      "epoch": 272.91,
+      "grad_norm": 14.364530563354492,
+      "learning_rate": 9.029029029029029e-07,
+      "loss": 0.3042,
+      "step": 90880
+    },
+    {
+      "epoch": 272.94,
+      "grad_norm": 18.584728240966797,
+      "learning_rate": 9.01901901901902e-07,
+      "loss": 0.3388,
+      "step": 90890
+    },
+    {
+      "epoch": 272.97,
+      "grad_norm": 24.936124801635742,
+      "learning_rate": 9.00900900900901e-07,
+      "loss": 0.2757,
+      "step": 90900
+    },
+    {
+      "epoch": 273.0,
+      "eval_accuracy": 0.8635,
+      "eval_loss": 0.5511497259140015,
+      "eval_runtime": 12.9013,
+      "eval_samples_per_second": 775.114,
+      "eval_steps_per_second": 3.1,
+      "step": 90909
+    },
+    {
+      "epoch": 273.0,
+      "grad_norm": 12.268407821655273,
+      "learning_rate": 8.998998998998999e-07,
+      "loss": 0.5775,
+      "step": 90910
+    },
+    {
+      "epoch": 273.03,
+      "grad_norm": 17.769210815429688,
+      "learning_rate": 8.988988988988989e-07,
+      "loss": 0.3274,
+      "step": 90920
+    },
+    {
+      "epoch": 273.06,
+      "grad_norm": 13.653009414672852,
+      "learning_rate": 8.978978978978979e-07,
+      "loss": 0.3106,
+      "step": 90930
+    },
+    {
+      "epoch": 273.09,
+      "grad_norm": 18.779882431030273,
+      "learning_rate": 8.968968968968969e-07,
+      "loss": 0.3162,
+      "step": 90940
+    },
+    {
+      "epoch": 273.12,
+      "grad_norm": 12.251233100891113,
+      "learning_rate": 8.958958958958958e-07,
+      "loss": 0.2796,
+      "step": 90950
+    },
+    {
+      "epoch": 273.15,
+      "grad_norm": 15.090911865234375,
+      "learning_rate": 8.948948948948951e-07,
+      "loss": 0.2823,
+      "step": 90960
+    },
+    {
+      "epoch": 273.18,
+      "grad_norm": 17.2552433013916,
+      "learning_rate": 8.93893893893894e-07,
+      "loss": 0.3344,
+      "step": 90970
+    },
+    {
+      "epoch": 273.21,
+      "grad_norm": 15.37354564666748,
+      "learning_rate": 8.92892892892893e-07,
+      "loss": 0.3352,
+      "step": 90980
+    },
+    {
+      "epoch": 273.24,
+      "grad_norm": 18.28791618347168,
+      "learning_rate": 8.91891891891892e-07,
+      "loss": 0.3794,
+      "step": 90990
+    },
+    {
+      "epoch": 273.27,
+      "grad_norm": 12.697564125061035,
+      "learning_rate": 8.90890890890891e-07,
+      "loss": 0.3351,
+      "step": 91000
+    },
+    {
+      "epoch": 273.3,
+      "grad_norm": 20.263154983520508,
+      "learning_rate": 8.8988988988989e-07,
+      "loss": 0.3315,
+      "step": 91010
+    },
+    {
+      "epoch": 273.33,
+      "grad_norm": 16.14650535583496,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 0.3244,
+      "step": 91020
+    },
+    {
+      "epoch": 273.36,
+      "grad_norm": 14.930522918701172,
+      "learning_rate": 8.87887887887888e-07,
+      "loss": 0.3518,
+      "step": 91030
+    },
+    {
+      "epoch": 273.39,
+      "grad_norm": 22.61672019958496,
+      "learning_rate": 8.86886886886887e-07,
+      "loss": 0.3778,
+      "step": 91040
+    },
+    {
+      "epoch": 273.42,
+      "grad_norm": 17.440141677856445,
+      "learning_rate": 8.858858858858859e-07,
+      "loss": 0.357,
+      "step": 91050
+    },
+    {
+      "epoch": 273.45,
+      "grad_norm": 9.335423469543457,
+      "learning_rate": 8.848848848848849e-07,
+      "loss": 0.2549,
+      "step": 91060
+    },
+    {
+      "epoch": 273.48,
+      "grad_norm": 17.331045150756836,
+      "learning_rate": 8.83883883883884e-07,
+      "loss": 0.3408,
+      "step": 91070
+    },
+    {
+      "epoch": 273.51,
+      "grad_norm": 23.115718841552734,
+      "learning_rate": 8.82882882882883e-07,
+      "loss": 0.3459,
+      "step": 91080
+    },
+    {
+      "epoch": 273.54,
+      "grad_norm": 9.009485244750977,
+      "learning_rate": 8.81881881881882e-07,
+      "loss": 0.3149,
+      "step": 91090
+    },
+    {
+      "epoch": 273.57,
+      "grad_norm": 13.758174896240234,
+      "learning_rate": 8.808808808808809e-07,
+      "loss": 0.3396,
+      "step": 91100
+    },
+    {
+      "epoch": 273.6,
+      "grad_norm": 14.7615327835083,
+      "learning_rate": 8.798798798798799e-07,
+      "loss": 0.3219,
+      "step": 91110
+    },
+    {
+      "epoch": 273.63,
+      "grad_norm": 23.169763565063477,
+      "learning_rate": 8.78878878878879e-07,
+      "loss": 0.3335,
+      "step": 91120
+    },
+    {
+      "epoch": 273.66,
+      "grad_norm": 13.73801326751709,
+      "learning_rate": 8.77877877877878e-07,
+      "loss": 0.3034,
+      "step": 91130
+    },
+    {
+      "epoch": 273.69,
+      "grad_norm": 15.88203239440918,
+      "learning_rate": 8.768768768768769e-07,
+      "loss": 0.3396,
+      "step": 91140
+    },
+    {
+      "epoch": 273.72,
+      "grad_norm": 15.930279731750488,
+      "learning_rate": 8.758758758758759e-07,
+      "loss": 0.3355,
+      "step": 91150
+    },
+    {
+      "epoch": 273.75,
+      "grad_norm": 20.308197021484375,
+      "learning_rate": 8.748748748748749e-07,
+      "loss": 0.3485,
+      "step": 91160
+    },
+    {
+      "epoch": 273.78,
+      "grad_norm": 11.442063331604004,
+      "learning_rate": 8.738738738738739e-07,
+      "loss": 0.3622,
+      "step": 91170
+    },
+    {
+      "epoch": 273.81,
+      "grad_norm": 18.331235885620117,
+      "learning_rate": 8.728728728728728e-07,
+      "loss": 0.2877,
+      "step": 91180
+    },
+    {
+      "epoch": 273.84,
+      "grad_norm": 22.612445831298828,
+      "learning_rate": 8.718718718718719e-07,
+      "loss": 0.2817,
+      "step": 91190
+    },
+    {
+      "epoch": 273.87,
+      "grad_norm": 14.113096237182617,
+      "learning_rate": 8.708708708708709e-07,
+      "loss": 0.3046,
+      "step": 91200
+    },
+    {
+      "epoch": 273.9,
+      "grad_norm": 15.800450325012207,
+      "learning_rate": 8.698698698698699e-07,
+      "loss": 0.3251,
+      "step": 91210
+    },
+    {
+      "epoch": 273.93,
+      "grad_norm": 12.551831245422363,
+      "learning_rate": 8.68868868868869e-07,
+      "loss": 0.2817,
+      "step": 91220
+    },
+    {
+      "epoch": 273.96,
+      "grad_norm": 13.25532054901123,
+      "learning_rate": 8.67867867867868e-07,
+      "loss": 0.2898,
+      "step": 91230
+    },
+    {
+      "epoch": 273.99,
+      "grad_norm": 14.129663467407227,
+      "learning_rate": 8.66866866866867e-07,
+      "loss": 0.372,
+      "step": 91240
+    },
+    {
+      "epoch": 274.0,
+      "eval_accuracy": 0.8633,
+      "eval_loss": 0.5518983602523804,
+      "eval_runtime": 12.7779,
+      "eval_samples_per_second": 782.604,
+      "eval_steps_per_second": 3.13,
+      "step": 91242
+    },
+    {
+      "epoch": 274.02,
+      "grad_norm": 18.735214233398438,
+      "learning_rate": 8.65865865865866e-07,
+      "loss": 0.3068,
+      "step": 91250
+    },
+    {
+      "epoch": 274.05,
+      "grad_norm": 18.59809112548828,
+      "learning_rate": 8.64864864864865e-07,
+      "loss": 0.3387,
+      "step": 91260
+    },
+    {
+      "epoch": 274.08,
+      "grad_norm": 12.077953338623047,
+      "learning_rate": 8.63863863863864e-07,
+      "loss": 0.3365,
+      "step": 91270
+    },
+    {
+      "epoch": 274.11,
+      "grad_norm": 11.23892593383789,
+      "learning_rate": 8.628628628628629e-07,
+      "loss": 0.2846,
+      "step": 91280
+    },
+    {
+      "epoch": 274.14,
+      "grad_norm": 13.634501457214355,
+      "learning_rate": 8.618618618618619e-07,
+      "loss": 0.3118,
+      "step": 91290
+    },
+    {
+      "epoch": 274.17,
+      "grad_norm": 23.556076049804688,
+      "learning_rate": 8.60860860860861e-07,
+      "loss": 0.3624,
+      "step": 91300
+    },
+    {
+      "epoch": 274.2,
+      "grad_norm": 22.51329231262207,
+      "learning_rate": 8.5985985985986e-07,
+      "loss": 0.2805,
+      "step": 91310
+    },
+    {
+      "epoch": 274.23,
+      "grad_norm": 15.054224014282227,
+      "learning_rate": 8.58858858858859e-07,
+      "loss": 0.3519,
+      "step": 91320
+    },
+    {
+      "epoch": 274.26,
+      "grad_norm": 15.54955768585205,
+      "learning_rate": 8.578578578578579e-07,
+      "loss": 0.2859,
+      "step": 91330
+    },
+    {
+      "epoch": 274.29,
+      "grad_norm": 12.030864715576172,
+      "learning_rate": 8.568568568568569e-07,
+      "loss": 0.2495,
+      "step": 91340
+    },
+    {
+      "epoch": 274.32,
+      "grad_norm": 16.658939361572266,
+      "learning_rate": 8.55855855855856e-07,
+      "loss": 0.4039,
+      "step": 91350
+    },
+    {
+      "epoch": 274.35,
+      "grad_norm": 11.693597793579102,
+      "learning_rate": 8.54854854854855e-07,
+      "loss": 0.3292,
+      "step": 91360
+    },
+    {
+      "epoch": 274.38,
+      "grad_norm": 13.52263355255127,
+      "learning_rate": 8.538538538538539e-07,
+      "loss": 0.3151,
+      "step": 91370
+    },
+    {
+      "epoch": 274.41,
+      "grad_norm": 15.198835372924805,
+      "learning_rate": 8.528528528528529e-07,
+      "loss": 0.3299,
+      "step": 91380
+    },
+    {
+      "epoch": 274.44,
+      "grad_norm": 18.716293334960938,
+      "learning_rate": 8.518518518518519e-07,
+      "loss": 0.3307,
+      "step": 91390
+    },
+    {
+      "epoch": 274.47,
+      "grad_norm": 23.904443740844727,
+      "learning_rate": 8.508508508508509e-07,
+      "loss": 0.3447,
+      "step": 91400
+    },
+    {
+      "epoch": 274.5,
+      "grad_norm": 27.3221378326416,
+      "learning_rate": 8.498498498498498e-07,
+      "loss": 0.3378,
+      "step": 91410
+    },
+    {
+      "epoch": 274.53,
+      "grad_norm": 13.758551597595215,
+      "learning_rate": 8.488488488488489e-07,
+      "loss": 0.3394,
+      "step": 91420
+    },
+    {
+      "epoch": 274.56,
+      "grad_norm": 14.256987571716309,
+      "learning_rate": 8.478478478478479e-07,
+      "loss": 0.2767,
+      "step": 91430
+    },
+    {
+      "epoch": 274.59,
+      "grad_norm": 15.280806541442871,
+      "learning_rate": 8.468468468468469e-07,
+      "loss": 0.3491,
+      "step": 91440
+    },
+    {
+      "epoch": 274.62,
+      "grad_norm": 32.0025749206543,
+      "learning_rate": 8.458458458458458e-07,
+      "loss": 0.3407,
+      "step": 91450
+    },
+    {
+      "epoch": 274.65,
+      "grad_norm": 13.056373596191406,
+      "learning_rate": 8.448448448448448e-07,
+      "loss": 0.3287,
+      "step": 91460
+    },
+    {
+      "epoch": 274.68,
+      "grad_norm": 14.12746524810791,
+      "learning_rate": 8.438438438438439e-07,
+      "loss": 0.3543,
+      "step": 91470
+    },
+    {
+      "epoch": 274.71,
+      "grad_norm": 19.862390518188477,
+      "learning_rate": 8.428428428428429e-07,
+      "loss": 0.4148,
+      "step": 91480
+    },
+    {
+      "epoch": 274.74,
+      "grad_norm": 18.470439910888672,
+      "learning_rate": 8.41841841841842e-07,
+      "loss": 0.3484,
+      "step": 91490
+    },
+    {
+      "epoch": 274.77,
+      "grad_norm": 15.435734748840332,
+      "learning_rate": 8.40840840840841e-07,
+      "loss": 0.291,
+      "step": 91500
+    },
+    {
+      "epoch": 274.8,
+      "grad_norm": 14.49168586730957,
+      "learning_rate": 8.398398398398399e-07,
+      "loss": 0.3155,
+      "step": 91510
+    },
+    {
+      "epoch": 274.83,
+      "grad_norm": 13.215459823608398,
+      "learning_rate": 8.388388388388389e-07,
+      "loss": 0.3061,
+      "step": 91520
+    },
+    {
+      "epoch": 274.86,
+      "grad_norm": 18.64816665649414,
+      "learning_rate": 8.37837837837838e-07,
+      "loss": 0.3466,
+      "step": 91530
+    },
+    {
+      "epoch": 274.89,
+      "grad_norm": 14.078080177307129,
+      "learning_rate": 8.36836836836837e-07,
+      "loss": 0.3273,
+      "step": 91540
+    },
+    {
+      "epoch": 274.92,
+      "grad_norm": 15.438807487487793,
+      "learning_rate": 8.358358358358359e-07,
+      "loss": 0.3732,
+      "step": 91550
+    },
+    {
+      "epoch": 274.95,
+      "grad_norm": 14.007704734802246,
+      "learning_rate": 8.348348348348349e-07,
+      "loss": 0.3128,
+      "step": 91560
+    },
+    {
+      "epoch": 274.98,
+      "grad_norm": 12.992770195007324,
+      "learning_rate": 8.338338338338339e-07,
+      "loss": 0.3262,
+      "step": 91570
+    },
+    {
+      "epoch": 275.0,
+      "eval_accuracy": 0.8641,
+      "eval_loss": 0.5502468943595886,
+      "eval_runtime": 12.7663,
+      "eval_samples_per_second": 783.313,
+      "eval_steps_per_second": 3.133,
+      "step": 91575
+    },
+    {
+      "epoch": 275.02,
+      "grad_norm": 12.626803398132324,
+      "learning_rate": 8.328328328328329e-07,
+      "loss": 0.3071,
+      "step": 91580
+    },
+    {
+      "epoch": 275.05,
+      "grad_norm": 21.2120418548584,
+      "learning_rate": 8.31831831831832e-07,
+      "loss": 0.3252,
+      "step": 91590
+    },
+    {
+      "epoch": 275.08,
+      "grad_norm": 10.949075698852539,
+      "learning_rate": 8.308308308308309e-07,
+      "loss": 0.3532,
+      "step": 91600
+    },
+    {
+      "epoch": 275.11,
+      "grad_norm": 13.762505531311035,
+      "learning_rate": 8.298298298298299e-07,
+      "loss": 0.3226,
+      "step": 91610
+    },
+    {
+      "epoch": 275.14,
+      "grad_norm": 15.778326988220215,
+      "learning_rate": 8.288288288288289e-07,
+      "loss": 0.354,
+      "step": 91620
+    },
+    {
+      "epoch": 275.17,
+      "grad_norm": 11.92486572265625,
+      "learning_rate": 8.278278278278279e-07,
+      "loss": 0.303,
+      "step": 91630
+    },
+    {
+      "epoch": 275.2,
+      "grad_norm": 12.388869285583496,
+      "learning_rate": 8.268268268268268e-07,
+      "loss": 0.308,
+      "step": 91640
+    },
+    {
+      "epoch": 275.23,
+      "grad_norm": 9.645380973815918,
+      "learning_rate": 8.258258258258259e-07,
+      "loss": 0.3254,
+      "step": 91650
+    },
+    {
+      "epoch": 275.26,
+      "grad_norm": 17.662324905395508,
+      "learning_rate": 8.248248248248249e-07,
+      "loss": 0.3249,
+      "step": 91660
+    },
+    {
+      "epoch": 275.29,
+      "grad_norm": 24.096435546875,
+      "learning_rate": 8.238238238238239e-07,
+      "loss": 0.3223,
+      "step": 91670
+    },
+    {
+      "epoch": 275.32,
+      "grad_norm": 17.233306884765625,
+      "learning_rate": 8.228228228228228e-07,
+      "loss": 0.3591,
+      "step": 91680
+    },
+    {
+      "epoch": 275.35,
+      "grad_norm": 12.89372444152832,
+      "learning_rate": 8.218218218218218e-07,
+      "loss": 0.3408,
+      "step": 91690
+    },
+    {
+      "epoch": 275.38,
+      "grad_norm": 17.64240837097168,
+      "learning_rate": 8.208208208208208e-07,
+      "loss": 0.3541,
+      "step": 91700
+    },
+    {
+      "epoch": 275.41,
+      "grad_norm": 14.329322814941406,
+      "learning_rate": 8.198198198198199e-07,
+      "loss": 0.3382,
+      "step": 91710
+    },
+    {
+      "epoch": 275.44,
+      "grad_norm": 14.23386001586914,
+      "learning_rate": 8.188188188188189e-07,
+      "loss": 0.3248,
+      "step": 91720
+    },
+    {
+      "epoch": 275.47,
+      "grad_norm": 12.045857429504395,
+      "learning_rate": 8.178178178178178e-07,
+      "loss": 0.395,
+      "step": 91730
+    },
+    {
+      "epoch": 275.5,
+      "grad_norm": 22.19173240661621,
+      "learning_rate": 8.168168168168168e-07,
+      "loss": 0.2961,
+      "step": 91740
+    },
+    {
+      "epoch": 275.53,
+      "grad_norm": 14.140740394592285,
+      "learning_rate": 8.158158158158158e-07,
+      "loss": 0.3499,
+      "step": 91750
+    },
+    {
+      "epoch": 275.56,
+      "grad_norm": 15.439727783203125,
+      "learning_rate": 8.14814814814815e-07,
+      "loss": 0.3321,
+      "step": 91760
+    },
+    {
+      "epoch": 275.59,
+      "grad_norm": 12.598774909973145,
+      "learning_rate": 8.13813813813814e-07,
+      "loss": 0.2993,
+      "step": 91770
+    },
+    {
+      "epoch": 275.62,
+      "grad_norm": 10.945440292358398,
+      "learning_rate": 8.128128128128129e-07,
+      "loss": 0.3169,
+      "step": 91780
+    },
+    {
+      "epoch": 275.65,
+      "grad_norm": 13.702495574951172,
+      "learning_rate": 8.118118118118119e-07,
+      "loss": 0.3271,
+      "step": 91790
+    },
+    {
+      "epoch": 275.68,
+      "grad_norm": 15.265227317810059,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.289,
+      "step": 91800
+    },
+    {
+      "epoch": 275.71,
+      "grad_norm": 16.282054901123047,
+      "learning_rate": 8.098098098098099e-07,
+      "loss": 0.3356,
+      "step": 91810
+    },
+    {
+      "epoch": 275.74,
+      "grad_norm": 13.679290771484375,
+      "learning_rate": 8.08808808808809e-07,
+      "loss": 0.3485,
+      "step": 91820
+    },
+    {
+      "epoch": 275.77,
+      "grad_norm": 21.71957778930664,
+      "learning_rate": 8.078078078078079e-07,
+      "loss": 0.3363,
+      "step": 91830
+    },
+    {
+      "epoch": 275.8,
+      "grad_norm": 19.62201499938965,
+      "learning_rate": 8.068068068068069e-07,
+      "loss": 0.3374,
+      "step": 91840
+    },
+    {
+      "epoch": 275.83,
+      "grad_norm": 22.832944869995117,
+      "learning_rate": 8.058058058058059e-07,
+      "loss": 0.2818,
+      "step": 91850
+    },
+    {
+      "epoch": 275.86,
+      "grad_norm": 10.706217765808105,
+      "learning_rate": 8.048048048048049e-07,
+      "loss": 0.3363,
+      "step": 91860
+    },
+    {
+      "epoch": 275.89,
+      "grad_norm": 14.631266593933105,
+      "learning_rate": 8.038038038038038e-07,
+      "loss": 0.2788,
+      "step": 91870
+    },
+    {
+      "epoch": 275.92,
+      "grad_norm": 12.055840492248535,
+      "learning_rate": 8.028028028028029e-07,
+      "loss": 0.3471,
+      "step": 91880
+    },
+    {
+      "epoch": 275.95,
+      "grad_norm": 21.967119216918945,
+      "learning_rate": 8.018018018018019e-07,
+      "loss": 0.3531,
+      "step": 91890
+    },
+    {
+      "epoch": 275.98,
+      "grad_norm": 13.03995418548584,
+      "learning_rate": 8.008008008008009e-07,
+      "loss": 0.2771,
+      "step": 91900
+    },
+    {
+      "epoch": 276.0,
+      "eval_accuracy": 0.8629,
+      "eval_loss": 0.5511269569396973,
+      "eval_runtime": 12.4955,
+      "eval_samples_per_second": 800.29,
+      "eval_steps_per_second": 3.201,
+      "step": 91908
+    },
+    {
+      "epoch": 276.01,
+      "grad_norm": 15.004878044128418,
+      "learning_rate": 7.997997997997998e-07,
+      "loss": 0.336,
+      "step": 91910
+    },
+    {
+      "epoch": 276.04,
+      "grad_norm": 14.659852027893066,
+      "learning_rate": 7.987987987987988e-07,
+      "loss": 0.3921,
+      "step": 91920
+    },
+    {
+      "epoch": 276.07,
+      "grad_norm": 20.157188415527344,
+      "learning_rate": 7.977977977977978e-07,
+      "loss": 0.3209,
+      "step": 91930
+    },
+    {
+      "epoch": 276.1,
+      "grad_norm": 18.346635818481445,
+      "learning_rate": 7.967967967967969e-07,
+      "loss": 0.3297,
+      "step": 91940
+    },
+    {
+      "epoch": 276.13,
+      "grad_norm": 17.01302719116211,
+      "learning_rate": 7.957957957957958e-07,
+      "loss": 0.3403,
+      "step": 91950
+    },
+    {
+      "epoch": 276.16,
+      "grad_norm": 17.55883026123047,
+      "learning_rate": 7.947947947947948e-07,
+      "loss": 0.3213,
+      "step": 91960
+    },
+    {
+      "epoch": 276.19,
+      "grad_norm": 22.240066528320312,
+      "learning_rate": 7.937937937937938e-07,
+      "loss": 0.3625,
+      "step": 91970
+    },
+    {
+      "epoch": 276.22,
+      "grad_norm": 12.544519424438477,
+      "learning_rate": 7.927927927927928e-07,
+      "loss": 0.3535,
+      "step": 91980
+    },
+    {
+      "epoch": 276.25,
+      "grad_norm": 13.400565147399902,
+      "learning_rate": 7.917917917917919e-07,
+      "loss": 0.3448,
+      "step": 91990
+    },
+    {
+      "epoch": 276.28,
+      "grad_norm": 14.514547348022461,
+      "learning_rate": 7.907907907907908e-07,
+      "loss": 0.3598,
+      "step": 92000
+    },
+    {
+      "epoch": 276.31,
+      "grad_norm": 20.177265167236328,
+      "learning_rate": 7.897897897897898e-07,
+      "loss": 0.3518,
+      "step": 92010
+    },
+    {
+      "epoch": 276.34,
+      "grad_norm": 12.411149978637695,
+      "learning_rate": 7.887887887887888e-07,
+      "loss": 0.3108,
+      "step": 92020
+    },
+    {
+      "epoch": 276.37,
+      "grad_norm": 14.413477897644043,
+      "learning_rate": 7.877877877877879e-07,
+      "loss": 0.3296,
+      "step": 92030
+    },
+    {
+      "epoch": 276.4,
+      "grad_norm": 22.919527053833008,
+      "learning_rate": 7.867867867867869e-07,
+      "loss": 0.3075,
+      "step": 92040
+    },
+    {
+      "epoch": 276.43,
+      "grad_norm": 12.77739429473877,
+      "learning_rate": 7.857857857857859e-07,
+      "loss": 0.3452,
+      "step": 92050
+    },
+    {
+      "epoch": 276.46,
+      "grad_norm": 13.35326862335205,
+      "learning_rate": 7.847847847847849e-07,
+      "loss": 0.352,
+      "step": 92060
+    },
+    {
+      "epoch": 276.49,
+      "grad_norm": 9.478172302246094,
+      "learning_rate": 7.837837837837839e-07,
+      "loss": 0.3376,
+      "step": 92070
+    },
+    {
+      "epoch": 276.52,
+      "grad_norm": 19.307538986206055,
+      "learning_rate": 7.827827827827829e-07,
+      "loss": 0.2777,
+      "step": 92080
+    },
+    {
+      "epoch": 276.55,
+      "grad_norm": 11.26225757598877,
+      "learning_rate": 7.817817817817819e-07,
+      "loss": 0.3734,
+      "step": 92090
+    },
+    {
+      "epoch": 276.58,
+      "grad_norm": 13.886388778686523,
+      "learning_rate": 7.807807807807808e-07,
+      "loss": 0.3581,
+      "step": 92100
+    },
+    {
+      "epoch": 276.61,
+      "grad_norm": 14.576263427734375,
+      "learning_rate": 7.797797797797799e-07,
+      "loss": 0.3117,
+      "step": 92110
+    },
+    {
+      "epoch": 276.64,
+      "grad_norm": 11.679303169250488,
+      "learning_rate": 7.787787787787789e-07,
+      "loss": 0.3526,
+      "step": 92120
+    },
+    {
+      "epoch": 276.67,
+      "grad_norm": 16.553123474121094,
+      "learning_rate": 7.777777777777779e-07,
+      "loss": 0.3662,
+      "step": 92130
+    },
+    {
+      "epoch": 276.7,
+      "grad_norm": 14.913849830627441,
+      "learning_rate": 7.767767767767768e-07,
+      "loss": 0.349,
+      "step": 92140
+    },
+    {
+      "epoch": 276.73,
+      "grad_norm": 15.175143241882324,
+      "learning_rate": 7.757757757757758e-07,
+      "loss": 0.3267,
+      "step": 92150
+    },
+    {
+      "epoch": 276.76,
+      "grad_norm": 16.324413299560547,
+      "learning_rate": 7.747747747747748e-07,
+      "loss": 0.3494,
+      "step": 92160
+    },
+    {
+      "epoch": 276.79,
+      "grad_norm": 20.552703857421875,
+      "learning_rate": 7.737737737737739e-07,
+      "loss": 0.3313,
+      "step": 92170
+    },
+    {
+      "epoch": 276.82,
+      "grad_norm": 14.763569831848145,
+      "learning_rate": 7.727727727727728e-07,
+      "loss": 0.335,
+      "step": 92180
+    },
+    {
+      "epoch": 276.85,
+      "grad_norm": 18.365568161010742,
+      "learning_rate": 7.717717717717718e-07,
+      "loss": 0.3225,
+      "step": 92190
+    },
+    {
+      "epoch": 276.88,
+      "grad_norm": 17.809627532958984,
+      "learning_rate": 7.707707707707708e-07,
+      "loss": 0.2936,
+      "step": 92200
+    },
+    {
+      "epoch": 276.91,
+      "grad_norm": 14.655138969421387,
+      "learning_rate": 7.697697697697698e-07,
+      "loss": 0.3252,
+      "step": 92210
+    },
+    {
+      "epoch": 276.94,
+      "grad_norm": 15.479802131652832,
+      "learning_rate": 7.687687687687688e-07,
+      "loss": 0.2999,
+      "step": 92220
+    },
+    {
+      "epoch": 276.97,
+      "grad_norm": 13.705923080444336,
+      "learning_rate": 7.677677677677678e-07,
+      "loss": 0.3175,
+      "step": 92230
+    },
+    {
+      "epoch": 277.0,
+      "grad_norm": 13.47728157043457,
+      "learning_rate": 7.667667667667668e-07,
+      "loss": 0.2907,
+      "step": 92240
+    },
+    {
+      "epoch": 277.0,
+      "eval_accuracy": 0.8627,
+      "eval_loss": 0.5511809587478638,
+      "eval_runtime": 12.4459,
+      "eval_samples_per_second": 803.477,
+      "eval_steps_per_second": 3.214,
+      "step": 92241
+    },
+    {
+      "epoch": 277.03,
+      "grad_norm": 11.667221069335938,
+      "learning_rate": 7.657657657657658e-07,
+      "loss": 0.3062,
+      "step": 92250
+    },
+    {
+      "epoch": 277.06,
+      "grad_norm": 18.63525390625,
+      "learning_rate": 7.647647647647648e-07,
+      "loss": 0.3408,
+      "step": 92260
+    },
+    {
+      "epoch": 277.09,
+      "grad_norm": 15.556292533874512,
+      "learning_rate": 7.637637637637637e-07,
+      "loss": 0.3463,
+      "step": 92270
+    },
+    {
+      "epoch": 277.12,
+      "grad_norm": 18.816028594970703,
+      "learning_rate": 7.627627627627627e-07,
+      "loss": 0.2632,
+      "step": 92280
+    },
+    {
+      "epoch": 277.15,
+      "grad_norm": 14.943379402160645,
+      "learning_rate": 7.617617617617619e-07,
+      "loss": 0.3276,
+      "step": 92290
+    },
+    {
+      "epoch": 277.18,
+      "grad_norm": 14.007725715637207,
+      "learning_rate": 7.607607607607609e-07,
+      "loss": 0.3019,
+      "step": 92300
+    },
+    {
+      "epoch": 277.21,
+      "grad_norm": 16.729591369628906,
+      "learning_rate": 7.597597597597599e-07,
+      "loss": 0.3326,
+      "step": 92310
+    },
+    {
+      "epoch": 277.24,
+      "grad_norm": 10.312609672546387,
+      "learning_rate": 7.587587587587588e-07,
+      "loss": 0.3289,
+      "step": 92320
+    },
+    {
+      "epoch": 277.27,
+      "grad_norm": 20.972097396850586,
+      "learning_rate": 7.577577577577578e-07,
+      "loss": 0.2897,
+      "step": 92330
+    },
+    {
+      "epoch": 277.3,
+      "grad_norm": 14.825124740600586,
+      "learning_rate": 7.567567567567569e-07,
+      "loss": 0.2941,
+      "step": 92340
+    },
+    {
+      "epoch": 277.33,
+      "grad_norm": 20.59674072265625,
+      "learning_rate": 7.557557557557559e-07,
+      "loss": 0.3421,
+      "step": 92350
+    },
+    {
+      "epoch": 277.36,
+      "grad_norm": 16.29591178894043,
+      "learning_rate": 7.547547547547549e-07,
+      "loss": 0.3696,
+      "step": 92360
+    },
+    {
+      "epoch": 277.39,
+      "grad_norm": 15.941308975219727,
+      "learning_rate": 7.537537537537538e-07,
+      "loss": 0.339,
+      "step": 92370
+    },
+    {
+      "epoch": 277.42,
+      "grad_norm": 15.867225646972656,
+      "learning_rate": 7.527527527527528e-07,
+      "loss": 0.3202,
+      "step": 92380
+    },
+    {
+      "epoch": 277.45,
+      "grad_norm": 16.362796783447266,
+      "learning_rate": 7.517517517517518e-07,
+      "loss": 0.3396,
+      "step": 92390
+    },
+    {
+      "epoch": 277.48,
+      "grad_norm": 19.394826889038086,
+      "learning_rate": 7.507507507507509e-07,
+      "loss": 0.2839,
+      "step": 92400
+    },
+    {
+      "epoch": 277.51,
+      "grad_norm": 23.144187927246094,
+      "learning_rate": 7.497497497497498e-07,
+      "loss": 0.3376,
+      "step": 92410
+    },
+    {
+      "epoch": 277.54,
+      "grad_norm": 20.861854553222656,
+      "learning_rate": 7.487487487487488e-07,
+      "loss": 0.3594,
+      "step": 92420
+    },
+    {
+      "epoch": 277.57,
+      "grad_norm": 28.429903030395508,
+      "learning_rate": 7.477477477477478e-07,
+      "loss": 0.3196,
+      "step": 92430
+    },
+    {
+      "epoch": 277.6,
+      "grad_norm": 21.1088924407959,
+      "learning_rate": 7.467467467467468e-07,
+      "loss": 0.3181,
+      "step": 92440
+    },
+    {
+      "epoch": 277.63,
+      "grad_norm": 13.960074424743652,
+      "learning_rate": 7.457457457457457e-07,
+      "loss": 0.3211,
+      "step": 92450
+    },
+    {
+      "epoch": 277.66,
+      "grad_norm": 14.858891487121582,
+      "learning_rate": 7.447447447447448e-07,
+      "loss": 0.3542,
+      "step": 92460
+    },
+    {
+      "epoch": 277.69,
+      "grad_norm": 12.159187316894531,
+      "learning_rate": 7.437437437437438e-07,
+      "loss": 0.331,
+      "step": 92470
+    },
+    {
+      "epoch": 277.72,
+      "grad_norm": 19.397197723388672,
+      "learning_rate": 7.427427427427428e-07,
+      "loss": 0.34,
+      "step": 92480
+    },
+    {
+      "epoch": 277.75,
+      "grad_norm": 28.440065383911133,
+      "learning_rate": 7.417417417417418e-07,
+      "loss": 0.3781,
+      "step": 92490
+    },
+    {
+      "epoch": 277.78,
+      "grad_norm": 22.974199295043945,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.3285,
+      "step": 92500
+    },
+    {
+      "epoch": 277.81,
+      "grad_norm": 15.243791580200195,
+      "learning_rate": 7.397397397397397e-07,
+      "loss": 0.3353,
+      "step": 92510
+    },
+    {
+      "epoch": 277.84,
+      "grad_norm": 8.77011489868164,
+      "learning_rate": 7.387387387387388e-07,
+      "loss": 0.2525,
+      "step": 92520
+    },
+    {
+      "epoch": 277.87,
+      "grad_norm": 19.961177825927734,
+      "learning_rate": 7.377377377377378e-07,
+      "loss": 0.2957,
+      "step": 92530
+    },
+    {
+      "epoch": 277.9,
+      "grad_norm": 13.038750648498535,
+      "learning_rate": 7.367367367367367e-07,
+      "loss": 0.2937,
+      "step": 92540
+    },
+    {
+      "epoch": 277.93,
+      "grad_norm": 14.883546829223633,
+      "learning_rate": 7.357357357357357e-07,
+      "loss": 0.3369,
+      "step": 92550
+    },
+    {
+      "epoch": 277.96,
+      "grad_norm": 11.387642860412598,
+      "learning_rate": 7.347347347347348e-07,
+      "loss": 0.2978,
+      "step": 92560
+    },
+    {
+      "epoch": 277.99,
+      "grad_norm": 13.90479850769043,
+      "learning_rate": 7.337337337337339e-07,
+      "loss": 0.3239,
+      "step": 92570
+    },
+    {
+      "epoch": 278.0,
+      "eval_accuracy": 0.8633,
+      "eval_loss": 0.5513429641723633,
+      "eval_runtime": 12.631,
+      "eval_samples_per_second": 791.701,
+      "eval_steps_per_second": 3.167,
+      "step": 92574
+    },
+    {
+      "epoch": 278.02,
+      "grad_norm": 16.46290397644043,
+      "learning_rate": 7.327327327327329e-07,
+      "loss": 0.3294,
+      "step": 92580
+    },
+    {
+      "epoch": 278.05,
+      "grad_norm": 15.394371032714844,
+      "learning_rate": 7.317317317317319e-07,
+      "loss": 0.3009,
+      "step": 92590
+    },
+    {
+      "epoch": 278.08,
+      "grad_norm": 14.995746612548828,
+      "learning_rate": 7.307307307307308e-07,
+      "loss": 0.3012,
+      "step": 92600
+    },
+    {
+      "epoch": 278.11,
+      "grad_norm": 15.097070693969727,
+      "learning_rate": 7.297297297297298e-07,
+      "loss": 0.3307,
+      "step": 92610
+    },
+    {
+      "epoch": 278.14,
+      "grad_norm": 15.02962875366211,
+      "learning_rate": 7.287287287287288e-07,
+      "loss": 0.3111,
+      "step": 92620
+    },
+    {
+      "epoch": 278.17,
+      "grad_norm": 19.245346069335938,
+      "learning_rate": 7.277277277277279e-07,
+      "loss": 0.2856,
+      "step": 92630
+    },
+    {
+      "epoch": 278.2,
+      "grad_norm": 14.440412521362305,
+      "learning_rate": 7.267267267267268e-07,
+      "loss": 0.3253,
+      "step": 92640
+    },
+    {
+      "epoch": 278.23,
+      "grad_norm": 20.759519577026367,
+      "learning_rate": 7.257257257257258e-07,
+      "loss": 0.3461,
+      "step": 92650
+    },
+    {
+      "epoch": 278.26,
+      "grad_norm": 18.45507049560547,
+      "learning_rate": 7.247247247247248e-07,
+      "loss": 0.3073,
+      "step": 92660
+    },
+    {
+      "epoch": 278.29,
+      "grad_norm": 17.24747085571289,
+      "learning_rate": 7.237237237237238e-07,
+      "loss": 0.3687,
+      "step": 92670
+    },
+    {
+      "epoch": 278.32,
+      "grad_norm": 13.567524909973145,
+      "learning_rate": 7.227227227227227e-07,
+      "loss": 0.3253,
+      "step": 92680
+    },
+    {
+      "epoch": 278.35,
+      "grad_norm": 12.105002403259277,
+      "learning_rate": 7.217217217217218e-07,
+      "loss": 0.3116,
+      "step": 92690
+    },
+    {
+      "epoch": 278.38,
+      "grad_norm": 14.260335922241211,
+      "learning_rate": 7.207207207207208e-07,
+      "loss": 0.3239,
+      "step": 92700
+    },
+    {
+      "epoch": 278.41,
+      "grad_norm": 11.4838228225708,
+      "learning_rate": 7.197197197197198e-07,
+      "loss": 0.3181,
+      "step": 92710
+    },
+    {
+      "epoch": 278.44,
+      "grad_norm": 19.57184410095215,
+      "learning_rate": 7.187187187187188e-07,
+      "loss": 0.3266,
+      "step": 92720
+    },
+    {
+      "epoch": 278.47,
+      "grad_norm": 14.970979690551758,
+      "learning_rate": 7.177177177177177e-07,
+      "loss": 0.2938,
+      "step": 92730
+    },
+    {
+      "epoch": 278.5,
+      "grad_norm": 19.05354881286621,
+      "learning_rate": 7.167167167167167e-07,
+      "loss": 0.3719,
+      "step": 92740
+    },
+    {
+      "epoch": 278.53,
+      "grad_norm": 22.050594329833984,
+      "learning_rate": 7.157157157157158e-07,
+      "loss": 0.3404,
+      "step": 92750
+    },
+    {
+      "epoch": 278.56,
+      "grad_norm": 14.784042358398438,
+      "learning_rate": 7.147147147147148e-07,
+      "loss": 0.286,
+      "step": 92760
+    },
+    {
+      "epoch": 278.59,
+      "grad_norm": 26.52376937866211,
+      "learning_rate": 7.137137137137137e-07,
+      "loss": 0.2959,
+      "step": 92770
+    },
+    {
+      "epoch": 278.62,
+      "grad_norm": 9.99184799194336,
+      "learning_rate": 7.127127127127127e-07,
+      "loss": 0.3053,
+      "step": 92780
+    },
+    {
+      "epoch": 278.65,
+      "grad_norm": 18.575584411621094,
+      "learning_rate": 7.117117117117117e-07,
+      "loss": 0.3467,
+      "step": 92790
+    },
+    {
+      "epoch": 278.68,
+      "grad_norm": 12.099264144897461,
+      "learning_rate": 7.107107107107107e-07,
+      "loss": 0.3174,
+      "step": 92800
+    },
+    {
+      "epoch": 278.71,
+      "grad_norm": 14.831299781799316,
+      "learning_rate": 7.097097097097097e-07,
+      "loss": 0.3682,
+      "step": 92810
+    },
+    {
+      "epoch": 278.74,
+      "grad_norm": 19.228424072265625,
+      "learning_rate": 7.087087087087087e-07,
+      "loss": 0.365,
+      "step": 92820
+    },
+    {
+      "epoch": 278.77,
+      "grad_norm": 15.119393348693848,
+      "learning_rate": 7.077077077077078e-07,
+      "loss": 0.3174,
+      "step": 92830
+    },
+    {
+      "epoch": 278.8,
+      "grad_norm": 23.911611557006836,
+      "learning_rate": 7.067067067067068e-07,
+      "loss": 0.3582,
+      "step": 92840
+    },
+    {
+      "epoch": 278.83,
+      "grad_norm": 11.42410659790039,
+      "learning_rate": 7.057057057057058e-07,
+      "loss": 0.3395,
+      "step": 92850
+    },
+    {
+      "epoch": 278.86,
+      "grad_norm": 19.268856048583984,
+      "learning_rate": 7.047047047047049e-07,
+      "loss": 0.2905,
+      "step": 92860
+    },
+    {
+      "epoch": 278.89,
+      "grad_norm": 25.38567352294922,
+      "learning_rate": 7.037037037037038e-07,
+      "loss": 0.3689,
+      "step": 92870
+    },
+    {
+      "epoch": 278.92,
+      "grad_norm": 13.112442970275879,
+      "learning_rate": 7.027027027027028e-07,
+      "loss": 0.3447,
+      "step": 92880
+    },
+    {
+      "epoch": 278.95,
+      "grad_norm": 11.552889823913574,
+      "learning_rate": 7.017017017017018e-07,
+      "loss": 0.2963,
+      "step": 92890
+    },
+    {
+      "epoch": 278.98,
+      "grad_norm": 17.18047332763672,
+      "learning_rate": 7.007007007007008e-07,
+      "loss": 0.2725,
+      "step": 92900
+    },
+    {
+      "epoch": 279.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5491702556610107,
+      "eval_runtime": 12.6983,
+      "eval_samples_per_second": 787.509,
+      "eval_steps_per_second": 3.15,
+      "step": 92907
+    },
+    {
+      "epoch": 279.01,
+      "grad_norm": 14.286107063293457,
+      "learning_rate": 6.996996996996997e-07,
+      "loss": 0.3371,
+      "step": 92910
+    },
+    {
+      "epoch": 279.04,
+      "grad_norm": 20.60944175720215,
+      "learning_rate": 6.986986986986988e-07,
+      "loss": 0.3126,
+      "step": 92920
+    },
+    {
+      "epoch": 279.07,
+      "grad_norm": 13.878959655761719,
+      "learning_rate": 6.976976976976978e-07,
+      "loss": 0.3162,
+      "step": 92930
+    },
+    {
+      "epoch": 279.1,
+      "grad_norm": 12.437176704406738,
+      "learning_rate": 6.966966966966968e-07,
+      "loss": 0.2991,
+      "step": 92940
+    },
+    {
+      "epoch": 279.13,
+      "grad_norm": 17.904760360717773,
+      "learning_rate": 6.956956956956957e-07,
+      "loss": 0.3647,
+      "step": 92950
+    },
+    {
+      "epoch": 279.16,
+      "grad_norm": 11.31718921661377,
+      "learning_rate": 6.946946946946947e-07,
+      "loss": 0.3091,
+      "step": 92960
+    },
+    {
+      "epoch": 279.19,
+      "grad_norm": 9.454315185546875,
+      "learning_rate": 6.936936936936937e-07,
+      "loss": 0.2986,
+      "step": 92970
+    },
+    {
+      "epoch": 279.22,
+      "grad_norm": 23.092798233032227,
+      "learning_rate": 6.926926926926928e-07,
+      "loss": 0.3313,
+      "step": 92980
+    },
+    {
+      "epoch": 279.25,
+      "grad_norm": 21.29292869567871,
+      "learning_rate": 6.916916916916918e-07,
+      "loss": 0.3522,
+      "step": 92990
+    },
+    {
+      "epoch": 279.28,
+      "grad_norm": 13.578611373901367,
+      "learning_rate": 6.906906906906907e-07,
+      "loss": 0.3339,
+      "step": 93000
+    },
+    {
+      "epoch": 279.31,
+      "grad_norm": 24.26846694946289,
+      "learning_rate": 6.896896896896897e-07,
+      "loss": 0.3354,
+      "step": 93010
+    },
+    {
+      "epoch": 279.34,
+      "grad_norm": 17.561893463134766,
+      "learning_rate": 6.886886886886887e-07,
+      "loss": 0.3267,
+      "step": 93020
+    },
+    {
+      "epoch": 279.37,
+      "grad_norm": 20.97976303100586,
+      "learning_rate": 6.876876876876877e-07,
+      "loss": 0.3396,
+      "step": 93030
+    },
+    {
+      "epoch": 279.4,
+      "grad_norm": 12.703837394714355,
+      "learning_rate": 6.866866866866867e-07,
+      "loss": 0.2901,
+      "step": 93040
+    },
+    {
+      "epoch": 279.43,
+      "grad_norm": 11.7171049118042,
+      "learning_rate": 6.856856856856857e-07,
+      "loss": 0.3449,
+      "step": 93050
+    },
+    {
+      "epoch": 279.46,
+      "grad_norm": 13.439323425292969,
+      "learning_rate": 6.846846846846847e-07,
+      "loss": 0.2982,
+      "step": 93060
+    },
+    {
+      "epoch": 279.49,
+      "grad_norm": 22.739604949951172,
+      "learning_rate": 6.836836836836837e-07,
+      "loss": 0.335,
+      "step": 93070
+    },
+    {
+      "epoch": 279.52,
+      "grad_norm": 17.70734214782715,
+      "learning_rate": 6.826826826826826e-07,
+      "loss": 0.3355,
+      "step": 93080
+    },
+    {
+      "epoch": 279.55,
+      "grad_norm": 16.00047492980957,
+      "learning_rate": 6.816816816816819e-07,
+      "loss": 0.3246,
+      "step": 93090
+    },
+    {
+      "epoch": 279.58,
+      "grad_norm": 18.70195770263672,
+      "learning_rate": 6.806806806806808e-07,
+      "loss": 0.3347,
+      "step": 93100
+    },
+    {
+      "epoch": 279.61,
+      "grad_norm": 17.2908878326416,
+      "learning_rate": 6.796796796796798e-07,
+      "loss": 0.3449,
+      "step": 93110
+    },
+    {
+      "epoch": 279.64,
+      "grad_norm": 13.979337692260742,
+      "learning_rate": 6.786786786786788e-07,
+      "loss": 0.3136,
+      "step": 93120
+    },
+    {
+      "epoch": 279.67,
+      "grad_norm": 14.769880294799805,
+      "learning_rate": 6.776776776776778e-07,
+      "loss": 0.3099,
+      "step": 93130
+    },
+    {
+      "epoch": 279.7,
+      "grad_norm": 14.013782501220703,
+      "learning_rate": 6.766766766766767e-07,
+      "loss": 0.2966,
+      "step": 93140
+    },
+    {
+      "epoch": 279.73,
+      "grad_norm": 14.062906265258789,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 0.3054,
+      "step": 93150
+    },
+    {
+      "epoch": 279.76,
+      "grad_norm": 15.284321784973145,
+      "learning_rate": 6.746746746746748e-07,
+      "loss": 0.3267,
+      "step": 93160
+    },
+    {
+      "epoch": 279.79,
+      "grad_norm": 11.912928581237793,
+      "learning_rate": 6.736736736736738e-07,
+      "loss": 0.3412,
+      "step": 93170
+    },
+    {
+      "epoch": 279.82,
+      "grad_norm": 19.985124588012695,
+      "learning_rate": 6.726726726726727e-07,
+      "loss": 0.3348,
+      "step": 93180
+    },
+    {
+      "epoch": 279.85,
+      "grad_norm": 8.725232124328613,
+      "learning_rate": 6.716716716716717e-07,
+      "loss": 0.2992,
+      "step": 93190
+    },
+    {
+      "epoch": 279.88,
+      "grad_norm": 17.65915298461914,
+      "learning_rate": 6.706706706706707e-07,
+      "loss": 0.3084,
+      "step": 93200
+    },
+    {
+      "epoch": 279.91,
+      "grad_norm": 14.579668045043945,
+      "learning_rate": 6.696696696696698e-07,
+      "loss": 0.2837,
+      "step": 93210
+    },
+    {
+      "epoch": 279.94,
+      "grad_norm": 19.889812469482422,
+      "learning_rate": 6.686686686686688e-07,
+      "loss": 0.326,
+      "step": 93220
+    },
+    {
+      "epoch": 279.97,
+      "grad_norm": 14.547228813171387,
+      "learning_rate": 6.676676676676677e-07,
+      "loss": 0.3275,
+      "step": 93230
+    },
+    {
+      "epoch": 280.0,
+      "grad_norm": 100.73583984375,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.3243,
+      "step": 93240
+    },
+    {
+      "epoch": 280.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.5504959225654602,
+      "eval_runtime": 12.7936,
+      "eval_samples_per_second": 781.638,
+      "eval_steps_per_second": 3.127,
+      "step": 93240
+    },
+    {
+      "epoch": 280.03,
+      "grad_norm": 20.667348861694336,
+      "learning_rate": 6.656656656656657e-07,
+      "loss": 0.3471,
+      "step": 93250
+    },
+    {
+      "epoch": 280.06,
+      "grad_norm": 16.233081817626953,
+      "learning_rate": 6.646646646646647e-07,
+      "loss": 0.3271,
+      "step": 93260
+    },
+    {
+      "epoch": 280.09,
+      "grad_norm": 15.62316608428955,
+      "learning_rate": 6.636636636636637e-07,
+      "loss": 0.3353,
+      "step": 93270
+    },
+    {
+      "epoch": 280.12,
+      "grad_norm": 19.024538040161133,
+      "learning_rate": 6.626626626626627e-07,
+      "loss": 0.3428,
+      "step": 93280
+    },
+    {
+      "epoch": 280.15,
+      "grad_norm": 10.936445236206055,
+      "learning_rate": 6.616616616616617e-07,
+      "loss": 0.2993,
+      "step": 93290
+    },
+    {
+      "epoch": 280.18,
+      "grad_norm": 21.69167137145996,
+      "learning_rate": 6.606606606606607e-07,
+      "loss": 0.325,
+      "step": 93300
+    },
+    {
+      "epoch": 280.21,
+      "grad_norm": 15.60485553741455,
+      "learning_rate": 6.596596596596596e-07,
+      "loss": 0.306,
+      "step": 93310
+    },
+    {
+      "epoch": 280.24,
+      "grad_norm": 17.066761016845703,
+      "learning_rate": 6.586586586586586e-07,
+      "loss": 0.3198,
+      "step": 93320
+    },
+    {
+      "epoch": 280.27,
+      "grad_norm": 21.2180233001709,
+      "learning_rate": 6.576576576576577e-07,
+      "loss": 0.3223,
+      "step": 93330
+    },
+    {
+      "epoch": 280.3,
+      "grad_norm": 15.150989532470703,
+      "learning_rate": 6.566566566566567e-07,
+      "loss": 0.3328,
+      "step": 93340
+    },
+    {
+      "epoch": 280.33,
+      "grad_norm": 14.779210090637207,
+      "learning_rate": 6.556556556556556e-07,
+      "loss": 0.3308,
+      "step": 93350
+    },
+    {
+      "epoch": 280.36,
+      "grad_norm": 12.310914039611816,
+      "learning_rate": 6.546546546546548e-07,
+      "loss": 0.3601,
+      "step": 93360
+    },
+    {
+      "epoch": 280.39,
+      "grad_norm": 16.876508712768555,
+      "learning_rate": 6.536536536536537e-07,
+      "loss": 0.3041,
+      "step": 93370
+    },
+    {
+      "epoch": 280.42,
+      "grad_norm": 11.565628051757812,
+      "learning_rate": 6.526526526526528e-07,
+      "loss": 0.3305,
+      "step": 93380
+    },
+    {
+      "epoch": 280.45,
+      "grad_norm": 21.497695922851562,
+      "learning_rate": 6.516516516516518e-07,
+      "loss": 0.3367,
+      "step": 93390
+    },
+    {
+      "epoch": 280.48,
+      "grad_norm": 9.66080379486084,
+      "learning_rate": 6.506506506506508e-07,
+      "loss": 0.3162,
+      "step": 93400
+    },
+    {
+      "epoch": 280.51,
+      "grad_norm": 12.400099754333496,
+      "learning_rate": 6.496496496496497e-07,
+      "loss": 0.3563,
+      "step": 93410
+    },
+    {
+      "epoch": 280.54,
+      "grad_norm": 19.505752563476562,
+      "learning_rate": 6.486486486486487e-07,
+      "loss": 0.3435,
+      "step": 93420
+    },
+    {
+      "epoch": 280.57,
+      "grad_norm": 22.753414154052734,
+      "learning_rate": 6.476476476476477e-07,
+      "loss": 0.3736,
+      "step": 93430
+    },
+    {
+      "epoch": 280.6,
+      "grad_norm": 15.251320838928223,
+      "learning_rate": 6.466466466466468e-07,
+      "loss": 0.3288,
+      "step": 93440
+    },
+    {
+      "epoch": 280.63,
+      "grad_norm": 15.568587303161621,
+      "learning_rate": 6.456456456456457e-07,
+      "loss": 0.303,
+      "step": 93450
+    },
+    {
+      "epoch": 280.66,
+      "grad_norm": 13.643176078796387,
+      "learning_rate": 6.446446446446447e-07,
+      "loss": 0.3008,
+      "step": 93460
+    },
+    {
+      "epoch": 280.69,
+      "grad_norm": 13.97219181060791,
+      "learning_rate": 6.436436436436437e-07,
+      "loss": 0.3011,
+      "step": 93470
+    },
+    {
+      "epoch": 280.72,
+      "grad_norm": 14.827840805053711,
+      "learning_rate": 6.426426426426427e-07,
+      "loss": 0.3116,
+      "step": 93480
+    },
+    {
+      "epoch": 280.75,
+      "grad_norm": 16.93248748779297,
+      "learning_rate": 6.416416416416417e-07,
+      "loss": 0.3476,
+      "step": 93490
+    },
+    {
+      "epoch": 280.78,
+      "grad_norm": 24.644289016723633,
+      "learning_rate": 6.406406406406407e-07,
+      "loss": 0.3271,
+      "step": 93500
+    },
+    {
+      "epoch": 280.81,
+      "grad_norm": 15.33893871307373,
+      "learning_rate": 6.396396396396397e-07,
+      "loss": 0.3253,
+      "step": 93510
+    },
+    {
+      "epoch": 280.84,
+      "grad_norm": 16.529796600341797,
+      "learning_rate": 6.386386386386387e-07,
+      "loss": 0.3563,
+      "step": 93520
+    },
+    {
+      "epoch": 280.87,
+      "grad_norm": 18.066003799438477,
+      "learning_rate": 6.376376376376377e-07,
+      "loss": 0.3193,
+      "step": 93530
+    },
+    {
+      "epoch": 280.9,
+      "grad_norm": 15.12454891204834,
+      "learning_rate": 6.366366366366366e-07,
+      "loss": 0.3072,
+      "step": 93540
+    },
+    {
+      "epoch": 280.93,
+      "grad_norm": 12.254704475402832,
+      "learning_rate": 6.356356356356356e-07,
+      "loss": 0.3026,
+      "step": 93550
+    },
+    {
+      "epoch": 280.96,
+      "grad_norm": 16.882230758666992,
+      "learning_rate": 6.346346346346347e-07,
+      "loss": 0.32,
+      "step": 93560
+    },
+    {
+      "epoch": 280.99,
+      "grad_norm": 18.799652099609375,
+      "learning_rate": 6.336336336336337e-07,
+      "loss": 0.3407,
+      "step": 93570
+    },
+    {
+      "epoch": 281.0,
+      "eval_accuracy": 0.8636,
+      "eval_loss": 0.5508015751838684,
+      "eval_runtime": 13.081,
+      "eval_samples_per_second": 764.466,
+      "eval_steps_per_second": 3.058,
+      "step": 93573
+    },
+    {
+      "epoch": 281.02,
+      "grad_norm": 14.940743446350098,
+      "learning_rate": 6.326326326326326e-07,
+      "loss": 0.3335,
+      "step": 93580
+    },
+    {
+      "epoch": 281.05,
+      "grad_norm": 22.657094955444336,
+      "learning_rate": 6.316316316316316e-07,
+      "loss": 0.3287,
+      "step": 93590
+    },
+    {
+      "epoch": 281.08,
+      "grad_norm": 15.750185012817383,
+      "learning_rate": 6.306306306306306e-07,
+      "loss": 0.3571,
+      "step": 93600
+    },
+    {
+      "epoch": 281.11,
+      "grad_norm": 14.329832077026367,
+      "learning_rate": 6.296296296296296e-07,
+      "loss": 0.3159,
+      "step": 93610
+    },
+    {
+      "epoch": 281.14,
+      "grad_norm": 26.950571060180664,
+      "learning_rate": 6.286286286286287e-07,
+      "loss": 0.3484,
+      "step": 93620
+    },
+    {
+      "epoch": 281.17,
+      "grad_norm": 15.880163192749023,
+      "learning_rate": 6.276276276276278e-07,
+      "loss": 0.3439,
+      "step": 93630
+    },
+    {
+      "epoch": 281.2,
+      "grad_norm": 21.715316772460938,
+      "learning_rate": 6.266266266266267e-07,
+      "loss": 0.308,
+      "step": 93640
+    },
+    {
+      "epoch": 281.23,
+      "grad_norm": 15.489581108093262,
+      "learning_rate": 6.256256256256257e-07,
+      "loss": 0.2884,
+      "step": 93650
+    },
+    {
+      "epoch": 281.26,
+      "grad_norm": 12.573485374450684,
+      "learning_rate": 6.246246246246246e-07,
+      "loss": 0.286,
+      "step": 93660
+    },
+    {
+      "epoch": 281.29,
+      "grad_norm": 8.714728355407715,
+      "learning_rate": 6.236236236236236e-07,
+      "loss": 0.3451,
+      "step": 93670
+    },
+    {
+      "epoch": 281.32,
+      "grad_norm": 18.92047691345215,
+      "learning_rate": 6.226226226226227e-07,
+      "loss": 0.323,
+      "step": 93680
+    },
+    {
+      "epoch": 281.35,
+      "grad_norm": 12.423954010009766,
+      "learning_rate": 6.216216216216217e-07,
+      "loss": 0.319,
+      "step": 93690
+    },
+    {
+      "epoch": 281.38,
+      "grad_norm": 13.857759475708008,
+      "learning_rate": 6.206206206206207e-07,
+      "loss": 0.3101,
+      "step": 93700
+    },
+    {
+      "epoch": 281.41,
+      "grad_norm": 16.481605529785156,
+      "learning_rate": 6.196196196196197e-07,
+      "loss": 0.3176,
+      "step": 93710
+    },
+    {
+      "epoch": 281.44,
+      "grad_norm": 18.58873748779297,
+      "learning_rate": 6.186186186186187e-07,
+      "loss": 0.3597,
+      "step": 93720
+    },
+    {
+      "epoch": 281.47,
+      "grad_norm": 14.239357948303223,
+      "learning_rate": 6.176176176176177e-07,
+      "loss": 0.3583,
+      "step": 93730
+    },
+    {
+      "epoch": 281.5,
+      "grad_norm": 14.787524223327637,
+      "learning_rate": 6.166166166166167e-07,
+      "loss": 0.3467,
+      "step": 93740
+    },
+    {
+      "epoch": 281.53,
+      "grad_norm": 15.881767272949219,
+      "learning_rate": 6.156156156156157e-07,
+      "loss": 0.3163,
+      "step": 93750
+    },
+    {
+      "epoch": 281.56,
+      "grad_norm": 17.96851348876953,
+      "learning_rate": 6.146146146146147e-07,
+      "loss": 0.2847,
+      "step": 93760
+    },
+    {
+      "epoch": 281.59,
+      "grad_norm": 10.404224395751953,
+      "learning_rate": 6.136136136136136e-07,
+      "loss": 0.3168,
+      "step": 93770
+    },
+    {
+      "epoch": 281.62,
+      "grad_norm": 14.23703384399414,
+      "learning_rate": 6.126126126126126e-07,
+      "loss": 0.3249,
+      "step": 93780
+    },
+    {
+      "epoch": 281.65,
+      "grad_norm": 16.447879791259766,
+      "learning_rate": 6.116116116116117e-07,
+      "loss": 0.3088,
+      "step": 93790
+    },
+    {
+      "epoch": 281.68,
+      "grad_norm": 18.41583824157715,
+      "learning_rate": 6.106106106106107e-07,
+      "loss": 0.3116,
+      "step": 93800
+    },
+    {
+      "epoch": 281.71,
+      "grad_norm": 17.694141387939453,
+      "learning_rate": 6.096096096096096e-07,
+      "loss": 0.3106,
+      "step": 93810
+    },
+    {
+      "epoch": 281.74,
+      "grad_norm": 12.786232948303223,
+      "learning_rate": 6.086086086086086e-07,
+      "loss": 0.3205,
+      "step": 93820
+    },
+    {
+      "epoch": 281.77,
+      "grad_norm": 10.402236938476562,
+      "learning_rate": 6.076076076076077e-07,
+      "loss": 0.3348,
+      "step": 93830
+    },
+    {
+      "epoch": 281.8,
+      "grad_norm": 14.408924102783203,
+      "learning_rate": 6.066066066066066e-07,
+      "loss": 0.3636,
+      "step": 93840
+    },
+    {
+      "epoch": 281.83,
+      "grad_norm": 12.40830135345459,
+      "learning_rate": 6.056056056056057e-07,
+      "loss": 0.3248,
+      "step": 93850
+    },
+    {
+      "epoch": 281.86,
+      "grad_norm": 20.661666870117188,
+      "learning_rate": 6.046046046046047e-07,
+      "loss": 0.3123,
+      "step": 93860
+    },
+    {
+      "epoch": 281.89,
+      "grad_norm": 18.561336517333984,
+      "learning_rate": 6.036036036036037e-07,
+      "loss": 0.2696,
+      "step": 93870
+    },
+    {
+      "epoch": 281.92,
+      "grad_norm": 11.23991870880127,
+      "learning_rate": 6.026026026026026e-07,
+      "loss": 0.2914,
+      "step": 93880
+    },
+    {
+      "epoch": 281.95,
+      "grad_norm": 16.677566528320312,
+      "learning_rate": 6.016016016016016e-07,
+      "loss": 0.3624,
+      "step": 93890
+    },
+    {
+      "epoch": 281.98,
+      "grad_norm": 18.102392196655273,
+      "learning_rate": 6.006006006006006e-07,
+      "loss": 0.3188,
+      "step": 93900
+    },
+    {
+      "epoch": 282.0,
+      "eval_accuracy": 0.8644,
+      "eval_loss": 0.5514655113220215,
+      "eval_runtime": 12.81,
+      "eval_samples_per_second": 780.638,
+      "eval_steps_per_second": 3.123,
+      "step": 93906
+    },
+    {
+      "epoch": 282.01,
+      "grad_norm": 28.10261344909668,
+      "learning_rate": 5.995995995995997e-07,
+      "loss": 0.5969,
+      "step": 93910
+    },
+    {
+      "epoch": 282.04,
+      "grad_norm": 13.727812767028809,
+      "learning_rate": 5.985985985985987e-07,
+      "loss": 0.3273,
+      "step": 93920
+    },
+    {
+      "epoch": 282.07,
+      "grad_norm": 16.154033660888672,
+      "learning_rate": 5.975975975975976e-07,
+      "loss": 0.3379,
+      "step": 93930
+    },
+    {
+      "epoch": 282.1,
+      "grad_norm": 23.370399475097656,
+      "learning_rate": 5.965965965965966e-07,
+      "loss": 0.3417,
+      "step": 93940
+    },
+    {
+      "epoch": 282.13,
+      "grad_norm": 14.242107391357422,
+      "learning_rate": 5.955955955955956e-07,
+      "loss": 0.2858,
+      "step": 93950
+    },
+    {
+      "epoch": 282.16,
+      "grad_norm": 26.32320785522461,
+      "learning_rate": 5.945945945945947e-07,
+      "loss": 0.3191,
+      "step": 93960
+    },
+    {
+      "epoch": 282.19,
+      "grad_norm": 30.024028778076172,
+      "learning_rate": 5.935935935935937e-07,
+      "loss": 0.2998,
+      "step": 93970
+    },
+    {
+      "epoch": 282.22,
+      "grad_norm": 19.347007751464844,
+      "learning_rate": 5.925925925925927e-07,
+      "loss": 0.3194,
+      "step": 93980
+    },
+    {
+      "epoch": 282.25,
+      "grad_norm": 8.81432819366455,
+      "learning_rate": 5.915915915915917e-07,
+      "loss": 0.3186,
+      "step": 93990
+    },
+    {
+      "epoch": 282.28,
+      "grad_norm": 14.406973838806152,
+      "learning_rate": 5.905905905905906e-07,
+      "loss": 0.3236,
+      "step": 94000
+    },
+    {
+      "epoch": 282.31,
+      "grad_norm": 23.851383209228516,
+      "learning_rate": 5.895895895895896e-07,
+      "loss": 0.3104,
+      "step": 94010
+    },
+    {
+      "epoch": 282.34,
+      "grad_norm": 17.779985427856445,
+      "learning_rate": 5.885885885885887e-07,
+      "loss": 0.3191,
+      "step": 94020
+    },
+    {
+      "epoch": 282.37,
+      "grad_norm": 15.324517250061035,
+      "learning_rate": 5.875875875875877e-07,
+      "loss": 0.3566,
+      "step": 94030
+    },
+    {
+      "epoch": 282.4,
+      "grad_norm": 13.028141975402832,
+      "learning_rate": 5.865865865865866e-07,
+      "loss": 0.3387,
+      "step": 94040
+    },
+    {
+      "epoch": 282.43,
+      "grad_norm": 7.575899124145508,
+      "learning_rate": 5.855855855855856e-07,
+      "loss": 0.3056,
+      "step": 94050
+    },
+    {
+      "epoch": 282.46,
+      "grad_norm": 11.889852523803711,
+      "learning_rate": 5.845845845845846e-07,
+      "loss": 0.3537,
+      "step": 94060
+    },
+    {
+      "epoch": 282.49,
+      "grad_norm": 13.052919387817383,
+      "learning_rate": 5.835835835835836e-07,
+      "loss": 0.3117,
+      "step": 94070
+    },
+    {
+      "epoch": 282.52,
+      "grad_norm": 12.842705726623535,
+      "learning_rate": 5.825825825825826e-07,
+      "loss": 0.3396,
+      "step": 94080
+    },
+    {
+      "epoch": 282.55,
+      "grad_norm": 12.911426544189453,
+      "learning_rate": 5.815815815815816e-07,
+      "loss": 0.3255,
+      "step": 94090
+    },
+    {
+      "epoch": 282.58,
+      "grad_norm": 15.252951622009277,
+      "learning_rate": 5.805805805805807e-07,
+      "loss": 0.3184,
+      "step": 94100
+    },
+    {
+      "epoch": 282.61,
+      "grad_norm": 14.325902938842773,
+      "learning_rate": 5.795795795795796e-07,
+      "loss": 0.3673,
+      "step": 94110
+    },
+    {
+      "epoch": 282.64,
+      "grad_norm": 16.862239837646484,
+      "learning_rate": 5.785785785785786e-07,
+      "loss": 0.3648,
+      "step": 94120
+    },
+    {
+      "epoch": 282.67,
+      "grad_norm": 12.80687427520752,
+      "learning_rate": 5.775775775775776e-07,
+      "loss": 0.2848,
+      "step": 94130
+    },
+    {
+      "epoch": 282.7,
+      "grad_norm": 22.02787208557129,
+      "learning_rate": 5.765765765765767e-07,
+      "loss": 0.3285,
+      "step": 94140
+    },
+    {
+      "epoch": 282.73,
+      "grad_norm": 9.854244232177734,
+      "learning_rate": 5.755755755755756e-07,
+      "loss": 0.3183,
+      "step": 94150
+    },
+    {
+      "epoch": 282.76,
+      "grad_norm": 20.77425193786621,
+      "learning_rate": 5.745745745745746e-07,
+      "loss": 0.3526,
+      "step": 94160
+    },
+    {
+      "epoch": 282.79,
+      "grad_norm": 12.556482315063477,
+      "learning_rate": 5.735735735735736e-07,
+      "loss": 0.3243,
+      "step": 94170
+    },
+    {
+      "epoch": 282.82,
+      "grad_norm": 21.712873458862305,
+      "learning_rate": 5.725725725725726e-07,
+      "loss": 0.2699,
+      "step": 94180
+    },
+    {
+      "epoch": 282.85,
+      "grad_norm": 16.457752227783203,
+      "learning_rate": 5.715715715715716e-07,
+      "loss": 0.3114,
+      "step": 94190
+    },
+    {
+      "epoch": 282.88,
+      "grad_norm": 14.076460838317871,
+      "learning_rate": 5.705705705705706e-07,
+      "loss": 0.2713,
+      "step": 94200
+    },
+    {
+      "epoch": 282.91,
+      "grad_norm": 15.028194427490234,
+      "learning_rate": 5.695695695695696e-07,
+      "loss": 0.3477,
+      "step": 94210
+    },
+    {
+      "epoch": 282.94,
+      "grad_norm": 18.480133056640625,
+      "learning_rate": 5.685685685685686e-07,
+      "loss": 0.3584,
+      "step": 94220
+    },
+    {
+      "epoch": 282.97,
+      "grad_norm": 17.163036346435547,
+      "learning_rate": 5.675675675675676e-07,
+      "loss": 0.3627,
+      "step": 94230
+    },
+    {
+      "epoch": 283.0,
+      "eval_accuracy": 0.8639,
+      "eval_loss": 0.5527836084365845,
+      "eval_runtime": 12.7864,
+      "eval_samples_per_second": 782.082,
+      "eval_steps_per_second": 3.128,
+      "step": 94239
+    },
+    {
+      "epoch": 283.0,
+      "grad_norm": 12.041302680969238,
+      "learning_rate": 5.665665665665666e-07,
+      "loss": 0.287,
+      "step": 94240
+    },
+    {
+      "epoch": 283.03,
+      "grad_norm": 16.460783004760742,
+      "learning_rate": 5.655655655655657e-07,
+      "loss": 0.3355,
+      "step": 94250
+    },
+    {
+      "epoch": 283.06,
+      "grad_norm": 30.21356201171875,
+      "learning_rate": 5.645645645645647e-07,
+      "loss": 0.3342,
+      "step": 94260
+    },
+    {
+      "epoch": 283.09,
+      "grad_norm": 14.283665657043457,
+      "learning_rate": 5.635635635635636e-07,
+      "loss": 0.3277,
+      "step": 94270
+    },
+    {
+      "epoch": 283.12,
+      "grad_norm": 17.427976608276367,
+      "learning_rate": 5.625625625625626e-07,
+      "loss": 0.3241,
+      "step": 94280
+    },
+    {
+      "epoch": 283.15,
+      "grad_norm": 17.976634979248047,
+      "learning_rate": 5.615615615615616e-07,
+      "loss": 0.3201,
+      "step": 94290
+    },
+    {
+      "epoch": 283.18,
+      "grad_norm": 14.567890167236328,
+      "learning_rate": 5.605605605605606e-07,
+      "loss": 0.2796,
+      "step": 94300
+    },
+    {
+      "epoch": 283.21,
+      "grad_norm": 21.419172286987305,
+      "learning_rate": 5.595595595595596e-07,
+      "loss": 0.3256,
+      "step": 94310
+    },
+    {
+      "epoch": 283.24,
+      "grad_norm": 16.18984031677246,
+      "learning_rate": 5.585585585585586e-07,
+      "loss": 0.3154,
+      "step": 94320
+    },
+    {
+      "epoch": 283.27,
+      "grad_norm": 22.039546966552734,
+      "learning_rate": 5.575575575575576e-07,
+      "loss": 0.3629,
+      "step": 94330
+    },
+    {
+      "epoch": 283.3,
+      "grad_norm": 22.113574981689453,
+      "learning_rate": 5.565565565565566e-07,
+      "loss": 0.3118,
+      "step": 94340
+    },
+    {
+      "epoch": 283.33,
+      "grad_norm": 9.19856071472168,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.3336,
+      "step": 94350
+    },
+    {
+      "epoch": 283.36,
+      "grad_norm": 23.878721237182617,
+      "learning_rate": 5.545545545545546e-07,
+      "loss": 0.3072,
+      "step": 94360
+    },
+    {
+      "epoch": 283.39,
+      "grad_norm": 24.39674949645996,
+      "learning_rate": 5.535535535535537e-07,
+      "loss": 0.2984,
+      "step": 94370
+    },
+    {
+      "epoch": 283.42,
+      "grad_norm": 27.68897819519043,
+      "learning_rate": 5.525525525525526e-07,
+      "loss": 0.2692,
+      "step": 94380
+    },
+    {
+      "epoch": 283.45,
+      "grad_norm": 25.7617130279541,
+      "learning_rate": 5.515515515515516e-07,
+      "loss": 0.3229,
+      "step": 94390
+    },
+    {
+      "epoch": 283.48,
+      "grad_norm": 9.298822402954102,
+      "learning_rate": 5.505505505505506e-07,
+      "loss": 0.289,
+      "step": 94400
+    },
+    {
+      "epoch": 283.51,
+      "grad_norm": 12.593667030334473,
+      "learning_rate": 5.495495495495496e-07,
+      "loss": 0.3046,
+      "step": 94410
+    },
+    {
+      "epoch": 283.54,
+      "grad_norm": 14.889416694641113,
+      "learning_rate": 5.485485485485486e-07,
+      "loss": 0.3218,
+      "step": 94420
+    },
+    {
+      "epoch": 283.57,
+      "grad_norm": 19.701887130737305,
+      "learning_rate": 5.475475475475476e-07,
+      "loss": 0.3085,
+      "step": 94430
+    },
+    {
+      "epoch": 283.6,
+      "grad_norm": 15.470821380615234,
+      "learning_rate": 5.465465465465466e-07,
+      "loss": 0.2785,
+      "step": 94440
+    },
+    {
+      "epoch": 283.63,
+      "grad_norm": 22.417850494384766,
+      "learning_rate": 5.455455455455456e-07,
+      "loss": 0.3425,
+      "step": 94450
+    },
+    {
+      "epoch": 283.66,
+      "grad_norm": 12.962305068969727,
+      "learning_rate": 5.445445445445446e-07,
+      "loss": 0.3026,
+      "step": 94460
+    },
+    {
+      "epoch": 283.69,
+      "grad_norm": 12.061803817749023,
+      "learning_rate": 5.435435435435435e-07,
+      "loss": 0.3313,
+      "step": 94470
+    },
+    {
+      "epoch": 283.72,
+      "grad_norm": 14.404483795166016,
+      "learning_rate": 5.425425425425425e-07,
+      "loss": 0.3179,
+      "step": 94480
+    },
+    {
+      "epoch": 283.75,
+      "grad_norm": 9.941262245178223,
+      "learning_rate": 5.415415415415416e-07,
+      "loss": 0.2987,
+      "step": 94490
+    },
+    {
+      "epoch": 283.78,
+      "grad_norm": 19.58894920349121,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.3249,
+      "step": 94500
+    },
+    {
+      "epoch": 283.81,
+      "grad_norm": 11.991168022155762,
+      "learning_rate": 5.395395395395396e-07,
+      "loss": 0.3121,
+      "step": 94510
+    },
+    {
+      "epoch": 283.84,
+      "grad_norm": 11.486674308776855,
+      "learning_rate": 5.385385385385386e-07,
+      "loss": 0.3489,
+      "step": 94520
+    },
+    {
+      "epoch": 283.87,
+      "grad_norm": 14.68239974975586,
+      "learning_rate": 5.375375375375376e-07,
+      "loss": 0.3062,
+      "step": 94530
+    },
+    {
+      "epoch": 283.9,
+      "grad_norm": 11.951064109802246,
+      "learning_rate": 5.365365365365365e-07,
+      "loss": 0.3194,
+      "step": 94540
+    },
+    {
+      "epoch": 283.93,
+      "grad_norm": 14.527950286865234,
+      "learning_rate": 5.355355355355356e-07,
+      "loss": 0.3256,
+      "step": 94550
+    },
+    {
+      "epoch": 283.96,
+      "grad_norm": 15.478323936462402,
+      "learning_rate": 5.345345345345346e-07,
+      "loss": 0.2845,
+      "step": 94560
+    },
+    {
+      "epoch": 283.99,
+      "grad_norm": 14.515961647033691,
+      "learning_rate": 5.335335335335336e-07,
+      "loss": 0.2879,
+      "step": 94570
+    },
+    {
+      "epoch": 284.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5517515540122986,
+      "eval_runtime": 12.8791,
+      "eval_samples_per_second": 776.451,
+      "eval_steps_per_second": 3.106,
+      "step": 94572
+    },
+    {
+      "epoch": 284.02,
+      "grad_norm": 16.581832885742188,
+      "learning_rate": 5.325325325325325e-07,
+      "loss": 0.3539,
+      "step": 94580
+    },
+    {
+      "epoch": 284.05,
+      "grad_norm": 13.50482177734375,
+      "learning_rate": 5.315315315315315e-07,
+      "loss": 0.3144,
+      "step": 94590
+    },
+    {
+      "epoch": 284.08,
+      "grad_norm": 18.097679138183594,
+      "learning_rate": 5.305305305305306e-07,
+      "loss": 0.3401,
+      "step": 94600
+    },
+    {
+      "epoch": 284.11,
+      "grad_norm": 15.857504844665527,
+      "learning_rate": 5.295295295295296e-07,
+      "loss": 0.344,
+      "step": 94610
+    },
+    {
+      "epoch": 284.14,
+      "grad_norm": 9.878615379333496,
+      "learning_rate": 5.285285285285286e-07,
+      "loss": 0.3042,
+      "step": 94620
+    },
+    {
+      "epoch": 284.17,
+      "grad_norm": 31.57953453063965,
+      "learning_rate": 5.275275275275276e-07,
+      "loss": 0.3491,
+      "step": 94630
+    },
+    {
+      "epoch": 284.2,
+      "grad_norm": 23.278371810913086,
+      "learning_rate": 5.265265265265266e-07,
+      "loss": 0.3342,
+      "step": 94640
+    },
+    {
+      "epoch": 284.23,
+      "grad_norm": 29.37053680419922,
+      "learning_rate": 5.255255255255255e-07,
+      "loss": 0.3299,
+      "step": 94650
+    },
+    {
+      "epoch": 284.26,
+      "grad_norm": 9.675458908081055,
+      "learning_rate": 5.245245245245246e-07,
+      "loss": 0.2746,
+      "step": 94660
+    },
+    {
+      "epoch": 284.29,
+      "grad_norm": 14.781416893005371,
+      "learning_rate": 5.235235235235236e-07,
+      "loss": 0.3509,
+      "step": 94670
+    },
+    {
+      "epoch": 284.32,
+      "grad_norm": 20.038114547729492,
+      "learning_rate": 5.225225225225226e-07,
+      "loss": 0.3298,
+      "step": 94680
+    },
+    {
+      "epoch": 284.35,
+      "grad_norm": 10.87721061706543,
+      "learning_rate": 5.215215215215216e-07,
+      "loss": 0.3245,
+      "step": 94690
+    },
+    {
+      "epoch": 284.38,
+      "grad_norm": 21.945188522338867,
+      "learning_rate": 5.205205205205205e-07,
+      "loss": 0.317,
+      "step": 94700
+    },
+    {
+      "epoch": 284.41,
+      "grad_norm": 11.684118270874023,
+      "learning_rate": 5.195195195195195e-07,
+      "loss": 0.3798,
+      "step": 94710
+    },
+    {
+      "epoch": 284.44,
+      "grad_norm": 10.704227447509766,
+      "learning_rate": 5.185185185185186e-07,
+      "loss": 0.3432,
+      "step": 94720
+    },
+    {
+      "epoch": 284.47,
+      "grad_norm": 18.698261260986328,
+      "learning_rate": 5.175175175175176e-07,
+      "loss": 0.3267,
+      "step": 94730
+    },
+    {
+      "epoch": 284.5,
+      "grad_norm": 19.02716636657715,
+      "learning_rate": 5.165165165165165e-07,
+      "loss": 0.2802,
+      "step": 94740
+    },
+    {
+      "epoch": 284.53,
+      "grad_norm": 12.124544143676758,
+      "learning_rate": 5.155155155155155e-07,
+      "loss": 0.2951,
+      "step": 94750
+    },
+    {
+      "epoch": 284.56,
+      "grad_norm": 12.476118087768555,
+      "learning_rate": 5.145145145145146e-07,
+      "loss": 0.3269,
+      "step": 94760
+    },
+    {
+      "epoch": 284.59,
+      "grad_norm": 18.51940155029297,
+      "learning_rate": 5.135135135135135e-07,
+      "loss": 0.3593,
+      "step": 94770
+    },
+    {
+      "epoch": 284.62,
+      "grad_norm": 18.55699348449707,
+      "learning_rate": 5.125125125125126e-07,
+      "loss": 0.3461,
+      "step": 94780
+    },
+    {
+      "epoch": 284.65,
+      "grad_norm": 19.993101119995117,
+      "learning_rate": 5.115115115115116e-07,
+      "loss": 0.3502,
+      "step": 94790
+    },
+    {
+      "epoch": 284.68,
+      "grad_norm": 11.905889511108398,
+      "learning_rate": 5.105105105105106e-07,
+      "loss": 0.2978,
+      "step": 94800
+    },
+    {
+      "epoch": 284.71,
+      "grad_norm": 17.218294143676758,
+      "learning_rate": 5.095095095095095e-07,
+      "loss": 0.3314,
+      "step": 94810
+    },
+    {
+      "epoch": 284.74,
+      "grad_norm": 16.78474235534668,
+      "learning_rate": 5.085085085085085e-07,
+      "loss": 0.3109,
+      "step": 94820
+    },
+    {
+      "epoch": 284.77,
+      "grad_norm": 15.883257865905762,
+      "learning_rate": 5.075075075075076e-07,
+      "loss": 0.3165,
+      "step": 94830
+    },
+    {
+      "epoch": 284.8,
+      "grad_norm": 19.85452651977539,
+      "learning_rate": 5.065065065065066e-07,
+      "loss": 0.3407,
+      "step": 94840
+    },
+    {
+      "epoch": 284.83,
+      "grad_norm": 23.443830490112305,
+      "learning_rate": 5.055055055055055e-07,
+      "loss": 0.3241,
+      "step": 94850
+    },
+    {
+      "epoch": 284.86,
+      "grad_norm": 22.895559310913086,
+      "learning_rate": 5.045045045045045e-07,
+      "loss": 0.3247,
+      "step": 94860
+    },
+    {
+      "epoch": 284.89,
+      "grad_norm": 15.995658874511719,
+      "learning_rate": 5.035035035035035e-07,
+      "loss": 0.3271,
+      "step": 94870
+    },
+    {
+      "epoch": 284.92,
+      "grad_norm": 21.718475341796875,
+      "learning_rate": 5.025025025025025e-07,
+      "loss": 0.2939,
+      "step": 94880
+    },
+    {
+      "epoch": 284.95,
+      "grad_norm": 13.846182823181152,
+      "learning_rate": 5.015015015015016e-07,
+      "loss": 0.3296,
+      "step": 94890
+    },
+    {
+      "epoch": 284.98,
+      "grad_norm": 14.658087730407715,
+      "learning_rate": 5.005005005005006e-07,
+      "loss": 0.288,
+      "step": 94900
+    },
+    {
+      "epoch": 285.0,
+      "eval_accuracy": 0.8633,
+      "eval_loss": 0.552438497543335,
+      "eval_runtime": 13.0025,
+      "eval_samples_per_second": 769.081,
+      "eval_steps_per_second": 3.076,
+      "step": 94905
+    },
+    {
+      "epoch": 285.02,
+      "grad_norm": 11.47140884399414,
+      "learning_rate": 4.994994994994996e-07,
+      "loss": 0.2578,
+      "step": 94910
+    },
+    {
+      "epoch": 285.05,
+      "grad_norm": 9.045263290405273,
+      "learning_rate": 4.984984984984986e-07,
+      "loss": 0.3388,
+      "step": 94920
+    },
+    {
+      "epoch": 285.08,
+      "grad_norm": 9.127370834350586,
+      "learning_rate": 4.974974974974975e-07,
+      "loss": 0.29,
+      "step": 94930
+    },
+    {
+      "epoch": 285.11,
+      "grad_norm": 13.511094093322754,
+      "learning_rate": 4.964964964964965e-07,
+      "loss": 0.3362,
+      "step": 94940
+    },
+    {
+      "epoch": 285.14,
+      "grad_norm": 16.988018035888672,
+      "learning_rate": 4.954954954954956e-07,
+      "loss": 0.2995,
+      "step": 94950
+    },
+    {
+      "epoch": 285.17,
+      "grad_norm": 26.761890411376953,
+      "learning_rate": 4.944944944944946e-07,
+      "loss": 0.3226,
+      "step": 94960
+    },
+    {
+      "epoch": 285.2,
+      "grad_norm": 18.649702072143555,
+      "learning_rate": 4.934934934934935e-07,
+      "loss": 0.2954,
+      "step": 94970
+    },
+    {
+      "epoch": 285.23,
+      "grad_norm": 13.229547500610352,
+      "learning_rate": 4.924924924924925e-07,
+      "loss": 0.3172,
+      "step": 94980
+    },
+    {
+      "epoch": 285.26,
+      "grad_norm": 19.608232498168945,
+      "learning_rate": 4.914914914914915e-07,
+      "loss": 0.3887,
+      "step": 94990
+    },
+    {
+      "epoch": 285.29,
+      "grad_norm": 11.743054389953613,
+      "learning_rate": 4.904904904904905e-07,
+      "loss": 0.3414,
+      "step": 95000
+    },
+    {
+      "epoch": 285.32,
+      "grad_norm": 14.482095718383789,
+      "learning_rate": 4.894894894894895e-07,
+      "loss": 0.3459,
+      "step": 95010
+    },
+    {
+      "epoch": 285.35,
+      "grad_norm": 20.459487915039062,
+      "learning_rate": 4.884884884884885e-07,
+      "loss": 0.2955,
+      "step": 95020
+    },
+    {
+      "epoch": 285.38,
+      "grad_norm": 20.43421173095703,
+      "learning_rate": 4.874874874874876e-07,
+      "loss": 0.3338,
+      "step": 95030
+    },
+    {
+      "epoch": 285.41,
+      "grad_norm": 10.059901237487793,
+      "learning_rate": 4.864864864864865e-07,
+      "loss": 0.3358,
+      "step": 95040
+    },
+    {
+      "epoch": 285.44,
+      "grad_norm": 18.449241638183594,
+      "learning_rate": 4.854854854854855e-07,
+      "loss": 0.3543,
+      "step": 95050
+    },
+    {
+      "epoch": 285.47,
+      "grad_norm": 19.0828800201416,
+      "learning_rate": 4.844844844844845e-07,
+      "loss": 0.334,
+      "step": 95060
+    },
+    {
+      "epoch": 285.5,
+      "grad_norm": 24.855257034301758,
+      "learning_rate": 4.834834834834836e-07,
+      "loss": 0.3198,
+      "step": 95070
+    },
+    {
+      "epoch": 285.53,
+      "grad_norm": 19.20456886291504,
+      "learning_rate": 4.824824824824825e-07,
+      "loss": 0.3687,
+      "step": 95080
+    },
+    {
+      "epoch": 285.56,
+      "grad_norm": 17.440031051635742,
+      "learning_rate": 4.814814814814815e-07,
+      "loss": 0.3065,
+      "step": 95090
+    },
+    {
+      "epoch": 285.59,
+      "grad_norm": 14.138654708862305,
+      "learning_rate": 4.804804804804805e-07,
+      "loss": 0.2885,
+      "step": 95100
+    },
+    {
+      "epoch": 285.62,
+      "grad_norm": 26.479724884033203,
+      "learning_rate": 4.794794794794795e-07,
+      "loss": 0.2792,
+      "step": 95110
+    },
+    {
+      "epoch": 285.65,
+      "grad_norm": 17.465452194213867,
+      "learning_rate": 4.784784784784786e-07,
+      "loss": 0.2853,
+      "step": 95120
+    },
+    {
+      "epoch": 285.68,
+      "grad_norm": 18.12849998474121,
+      "learning_rate": 4.774774774774775e-07,
+      "loss": 0.3352,
+      "step": 95130
+    },
+    {
+      "epoch": 285.71,
+      "grad_norm": 18.491870880126953,
+      "learning_rate": 4.764764764764765e-07,
+      "loss": 0.3353,
+      "step": 95140
+    },
+    {
+      "epoch": 285.74,
+      "grad_norm": 16.067407608032227,
+      "learning_rate": 4.754754754754755e-07,
+      "loss": 0.3214,
+      "step": 95150
+    },
+    {
+      "epoch": 285.77,
+      "grad_norm": 19.190481185913086,
+      "learning_rate": 4.744744744744745e-07,
+      "loss": 0.295,
+      "step": 95160
+    },
+    {
+      "epoch": 285.8,
+      "grad_norm": 19.59402084350586,
+      "learning_rate": 4.7347347347347354e-07,
+      "loss": 0.3238,
+      "step": 95170
+    },
+    {
+      "epoch": 285.83,
+      "grad_norm": 29.57146453857422,
+      "learning_rate": 4.7247247247247256e-07,
+      "loss": 0.2978,
+      "step": 95180
+    },
+    {
+      "epoch": 285.86,
+      "grad_norm": 14.194268226623535,
+      "learning_rate": 4.714714714714715e-07,
+      "loss": 0.2965,
+      "step": 95190
+    },
+    {
+      "epoch": 285.89,
+      "grad_norm": 25.47725486755371,
+      "learning_rate": 4.7047047047047054e-07,
+      "loss": 0.2981,
+      "step": 95200
+    },
+    {
+      "epoch": 285.92,
+      "grad_norm": 17.67198371887207,
+      "learning_rate": 4.694694694694695e-07,
+      "loss": 0.3805,
+      "step": 95210
+    },
+    {
+      "epoch": 285.95,
+      "grad_norm": 12.546442031860352,
+      "learning_rate": 4.684684684684685e-07,
+      "loss": 0.3207,
+      "step": 95220
+    },
+    {
+      "epoch": 285.98,
+      "grad_norm": 14.795281410217285,
+      "learning_rate": 4.674674674674675e-07,
+      "loss": 0.3104,
+      "step": 95230
+    },
+    {
+      "epoch": 286.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5516053438186646,
+      "eval_runtime": 12.9914,
+      "eval_samples_per_second": 769.741,
+      "eval_steps_per_second": 3.079,
+      "step": 95238
+    },
+    {
+      "epoch": 286.01,
+      "grad_norm": 13.881553649902344,
+      "learning_rate": 4.664664664664665e-07,
+      "loss": 0.2964,
+      "step": 95240
+    },
+    {
+      "epoch": 286.04,
+      "grad_norm": 19.25999641418457,
+      "learning_rate": 4.654654654654655e-07,
+      "loss": 0.2919,
+      "step": 95250
+    },
+    {
+      "epoch": 286.07,
+      "grad_norm": 14.622572898864746,
+      "learning_rate": 4.644644644644645e-07,
+      "loss": 0.3222,
+      "step": 95260
+    },
+    {
+      "epoch": 286.1,
+      "grad_norm": 13.258309364318848,
+      "learning_rate": 4.6346346346346346e-07,
+      "loss": 0.321,
+      "step": 95270
+    },
+    {
+      "epoch": 286.13,
+      "grad_norm": 17.333141326904297,
+      "learning_rate": 4.624624624624625e-07,
+      "loss": 0.2629,
+      "step": 95280
+    },
+    {
+      "epoch": 286.16,
+      "grad_norm": 13.932476997375488,
+      "learning_rate": 4.6146146146146144e-07,
+      "loss": 0.3553,
+      "step": 95290
+    },
+    {
+      "epoch": 286.19,
+      "grad_norm": 15.53055191040039,
+      "learning_rate": 4.604604604604605e-07,
+      "loss": 0.2841,
+      "step": 95300
+    },
+    {
+      "epoch": 286.22,
+      "grad_norm": 18.521644592285156,
+      "learning_rate": 4.5945945945945953e-07,
+      "loss": 0.3744,
+      "step": 95310
+    },
+    {
+      "epoch": 286.25,
+      "grad_norm": 12.427465438842773,
+      "learning_rate": 4.584584584584585e-07,
+      "loss": 0.3216,
+      "step": 95320
+    },
+    {
+      "epoch": 286.28,
+      "grad_norm": 24.99172592163086,
+      "learning_rate": 4.574574574574575e-07,
+      "loss": 0.3383,
+      "step": 95330
+    },
+    {
+      "epoch": 286.31,
+      "grad_norm": 11.157258033752441,
+      "learning_rate": 4.564564564564565e-07,
+      "loss": 0.319,
+      "step": 95340
+    },
+    {
+      "epoch": 286.34,
+      "grad_norm": 14.750675201416016,
+      "learning_rate": 4.554554554554555e-07,
+      "loss": 0.2985,
+      "step": 95350
+    },
+    {
+      "epoch": 286.37,
+      "grad_norm": 17.175037384033203,
+      "learning_rate": 4.544544544544545e-07,
+      "loss": 0.3301,
+      "step": 95360
+    },
+    {
+      "epoch": 286.4,
+      "grad_norm": 12.235640525817871,
+      "learning_rate": 4.534534534534535e-07,
+      "loss": 0.35,
+      "step": 95370
+    },
+    {
+      "epoch": 286.43,
+      "grad_norm": 19.873987197875977,
+      "learning_rate": 4.524524524524525e-07,
+      "loss": 0.3883,
+      "step": 95380
+    },
+    {
+      "epoch": 286.46,
+      "grad_norm": 18.04424285888672,
+      "learning_rate": 4.5145145145145147e-07,
+      "loss": 0.327,
+      "step": 95390
+    },
+    {
+      "epoch": 286.49,
+      "grad_norm": 17.59967803955078,
+      "learning_rate": 4.504504504504505e-07,
+      "loss": 0.2778,
+      "step": 95400
+    },
+    {
+      "epoch": 286.52,
+      "grad_norm": 20.002147674560547,
+      "learning_rate": 4.4944944944944945e-07,
+      "loss": 0.3638,
+      "step": 95410
+    },
+    {
+      "epoch": 286.55,
+      "grad_norm": 15.550737380981445,
+      "learning_rate": 4.4844844844844847e-07,
+      "loss": 0.2835,
+      "step": 95420
+    },
+    {
+      "epoch": 286.58,
+      "grad_norm": 23.610403060913086,
+      "learning_rate": 4.4744744744744754e-07,
+      "loss": 0.3859,
+      "step": 95430
+    },
+    {
+      "epoch": 286.61,
+      "grad_norm": 12.391138076782227,
+      "learning_rate": 4.464464464464465e-07,
+      "loss": 0.3269,
+      "step": 95440
+    },
+    {
+      "epoch": 286.64,
+      "grad_norm": 11.270466804504395,
+      "learning_rate": 4.454454454454455e-07,
+      "loss": 0.2732,
+      "step": 95450
+    },
+    {
+      "epoch": 286.67,
+      "grad_norm": 12.160772323608398,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 0.3185,
+      "step": 95460
+    },
+    {
+      "epoch": 286.7,
+      "grad_norm": 16.61357307434082,
+      "learning_rate": 4.434434434434435e-07,
+      "loss": 0.3329,
+      "step": 95470
+    },
+    {
+      "epoch": 286.73,
+      "grad_norm": 16.967079162597656,
+      "learning_rate": 4.4244244244244247e-07,
+      "loss": 0.3188,
+      "step": 95480
+    },
+    {
+      "epoch": 286.76,
+      "grad_norm": 11.631878852844238,
+      "learning_rate": 4.414414414414415e-07,
+      "loss": 0.3102,
+      "step": 95490
+    },
+    {
+      "epoch": 286.79,
+      "grad_norm": 12.391975402832031,
+      "learning_rate": 4.4044044044044046e-07,
+      "loss": 0.3061,
+      "step": 95500
+    },
+    {
+      "epoch": 286.82,
+      "grad_norm": 25.9937744140625,
+      "learning_rate": 4.394394394394395e-07,
+      "loss": 0.3706,
+      "step": 95510
+    },
+    {
+      "epoch": 286.85,
+      "grad_norm": 15.875472068786621,
+      "learning_rate": 4.3843843843843844e-07,
+      "loss": 0.2943,
+      "step": 95520
+    },
+    {
+      "epoch": 286.88,
+      "grad_norm": 13.135638236999512,
+      "learning_rate": 4.3743743743743746e-07,
+      "loss": 0.3191,
+      "step": 95530
+    },
+    {
+      "epoch": 286.91,
+      "grad_norm": 19.847158432006836,
+      "learning_rate": 4.364364364364364e-07,
+      "loss": 0.3516,
+      "step": 95540
+    },
+    {
+      "epoch": 286.94,
+      "grad_norm": 15.069009780883789,
+      "learning_rate": 4.3543543543543544e-07,
+      "loss": 0.3045,
+      "step": 95550
+    },
+    {
+      "epoch": 286.97,
+      "grad_norm": 16.61595916748047,
+      "learning_rate": 4.344344344344345e-07,
+      "loss": 0.3652,
+      "step": 95560
+    },
+    {
+      "epoch": 287.0,
+      "grad_norm": 15.168715476989746,
+      "learning_rate": 4.334334334334335e-07,
+      "loss": 0.2963,
+      "step": 95570
+    },
+    {
+      "epoch": 287.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.5508705973625183,
+      "eval_runtime": 13.0869,
+      "eval_samples_per_second": 764.125,
+      "eval_steps_per_second": 3.056,
+      "step": 95571
+    },
+    {
+      "epoch": 287.03,
+      "grad_norm": 24.199548721313477,
+      "learning_rate": 4.324324324324325e-07,
+      "loss": 0.3579,
+      "step": 95580
+    },
+    {
+      "epoch": 287.06,
+      "grad_norm": 11.737783432006836,
+      "learning_rate": 4.3143143143143146e-07,
+      "loss": 0.3151,
+      "step": 95590
+    },
+    {
+      "epoch": 287.09,
+      "grad_norm": 19.371213912963867,
+      "learning_rate": 4.304304304304305e-07,
+      "loss": 0.2748,
+      "step": 95600
+    },
+    {
+      "epoch": 287.12,
+      "grad_norm": 13.646231651306152,
+      "learning_rate": 4.294294294294295e-07,
+      "loss": 0.3163,
+      "step": 95610
+    },
+    {
+      "epoch": 287.15,
+      "grad_norm": 14.932293891906738,
+      "learning_rate": 4.2842842842842846e-07,
+      "loss": 0.3251,
+      "step": 95620
+    },
+    {
+      "epoch": 287.18,
+      "grad_norm": 17.453428268432617,
+      "learning_rate": 4.274274274274275e-07,
+      "loss": 0.291,
+      "step": 95630
+    },
+    {
+      "epoch": 287.21,
+      "grad_norm": 9.312527656555176,
+      "learning_rate": 4.2642642642642645e-07,
+      "loss": 0.286,
+      "step": 95640
+    },
+    {
+      "epoch": 287.24,
+      "grad_norm": 18.840839385986328,
+      "learning_rate": 4.2542542542542547e-07,
+      "loss": 0.3543,
+      "step": 95650
+    },
+    {
+      "epoch": 287.27,
+      "grad_norm": 14.779626846313477,
+      "learning_rate": 4.2442442442442443e-07,
+      "loss": 0.3497,
+      "step": 95660
+    },
+    {
+      "epoch": 287.3,
+      "grad_norm": 13.632381439208984,
+      "learning_rate": 4.2342342342342345e-07,
+      "loss": 0.349,
+      "step": 95670
+    },
+    {
+      "epoch": 287.33,
+      "grad_norm": 14.272237777709961,
+      "learning_rate": 4.224224224224224e-07,
+      "loss": 0.3176,
+      "step": 95680
+    },
+    {
+      "epoch": 287.36,
+      "grad_norm": 19.885513305664062,
+      "learning_rate": 4.2142142142142143e-07,
+      "loss": 0.2692,
+      "step": 95690
+    },
+    {
+      "epoch": 287.39,
+      "grad_norm": 17.83985710144043,
+      "learning_rate": 4.204204204204205e-07,
+      "loss": 0.3099,
+      "step": 95700
+    },
+    {
+      "epoch": 287.42,
+      "grad_norm": 22.414592742919922,
+      "learning_rate": 4.1941941941941947e-07,
+      "loss": 0.3227,
+      "step": 95710
+    },
+    {
+      "epoch": 287.45,
+      "grad_norm": 16.904897689819336,
+      "learning_rate": 4.184184184184185e-07,
+      "loss": 0.3144,
+      "step": 95720
+    },
+    {
+      "epoch": 287.48,
+      "grad_norm": 16.848430633544922,
+      "learning_rate": 4.1741741741741745e-07,
+      "loss": 0.3602,
+      "step": 95730
+    },
+    {
+      "epoch": 287.51,
+      "grad_norm": 22.242162704467773,
+      "learning_rate": 4.1641641641641647e-07,
+      "loss": 0.3185,
+      "step": 95740
+    },
+    {
+      "epoch": 287.54,
+      "grad_norm": 16.88922691345215,
+      "learning_rate": 4.1541541541541544e-07,
+      "loss": 0.3283,
+      "step": 95750
+    },
+    {
+      "epoch": 287.57,
+      "grad_norm": 19.56884002685547,
+      "learning_rate": 4.1441441441441446e-07,
+      "loss": 0.3141,
+      "step": 95760
+    },
+    {
+      "epoch": 287.6,
+      "grad_norm": 27.038545608520508,
+      "learning_rate": 4.134134134134134e-07,
+      "loss": 0.3288,
+      "step": 95770
+    },
+    {
+      "epoch": 287.63,
+      "grad_norm": 20.162240982055664,
+      "learning_rate": 4.1241241241241244e-07,
+      "loss": 0.3581,
+      "step": 95780
+    },
+    {
+      "epoch": 287.66,
+      "grad_norm": 29.0164852142334,
+      "learning_rate": 4.114114114114114e-07,
+      "loss": 0.2959,
+      "step": 95790
+    },
+    {
+      "epoch": 287.69,
+      "grad_norm": 22.30294418334961,
+      "learning_rate": 4.104104104104104e-07,
+      "loss": 0.3121,
+      "step": 95800
+    },
+    {
+      "epoch": 287.72,
+      "grad_norm": 20.47224998474121,
+      "learning_rate": 4.0940940940940944e-07,
+      "loss": 0.2985,
+      "step": 95810
+    },
+    {
+      "epoch": 287.75,
+      "grad_norm": 19.868282318115234,
+      "learning_rate": 4.084084084084084e-07,
+      "loss": 0.3408,
+      "step": 95820
+    },
+    {
+      "epoch": 287.78,
+      "grad_norm": 21.673015594482422,
+      "learning_rate": 4.074074074074075e-07,
+      "loss": 0.3011,
+      "step": 95830
+    },
+    {
+      "epoch": 287.81,
+      "grad_norm": 12.04843521118164,
+      "learning_rate": 4.0640640640640644e-07,
+      "loss": 0.3142,
+      "step": 95840
+    },
+    {
+      "epoch": 287.84,
+      "grad_norm": 21.418336868286133,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 0.3503,
+      "step": 95850
+    },
+    {
+      "epoch": 287.87,
+      "grad_norm": 10.616264343261719,
+      "learning_rate": 4.044044044044045e-07,
+      "loss": 0.3088,
+      "step": 95860
+    },
+    {
+      "epoch": 287.9,
+      "grad_norm": 14.000981330871582,
+      "learning_rate": 4.0340340340340345e-07,
+      "loss": 0.3425,
+      "step": 95870
+    },
+    {
+      "epoch": 287.93,
+      "grad_norm": 15.844428062438965,
+      "learning_rate": 4.0240240240240246e-07,
+      "loss": 0.3076,
+      "step": 95880
+    },
+    {
+      "epoch": 287.96,
+      "grad_norm": 16.5364990234375,
+      "learning_rate": 4.0140140140140143e-07,
+      "loss": 0.3226,
+      "step": 95890
+    },
+    {
+      "epoch": 287.99,
+      "grad_norm": 15.378037452697754,
+      "learning_rate": 4.0040040040040045e-07,
+      "loss": 0.348,
+      "step": 95900
+    },
+    {
+      "epoch": 288.0,
+      "eval_accuracy": 0.8647,
+      "eval_loss": 0.5511307120323181,
+      "eval_runtime": 12.9523,
+      "eval_samples_per_second": 772.066,
+      "eval_steps_per_second": 3.088,
+      "step": 95904
+    },
+    {
+      "epoch": 288.02,
+      "grad_norm": 13.313666343688965,
+      "learning_rate": 3.993993993993994e-07,
+      "loss": 0.3908,
+      "step": 95910
+    },
+    {
+      "epoch": 288.05,
+      "grad_norm": 20.2457218170166,
+      "learning_rate": 3.9839839839839843e-07,
+      "loss": 0.2977,
+      "step": 95920
+    },
+    {
+      "epoch": 288.08,
+      "grad_norm": 14.275227546691895,
+      "learning_rate": 3.973973973973974e-07,
+      "loss": 0.3186,
+      "step": 95930
+    },
+    {
+      "epoch": 288.11,
+      "grad_norm": 26.77968406677246,
+      "learning_rate": 3.963963963963964e-07,
+      "loss": 0.3081,
+      "step": 95940
+    },
+    {
+      "epoch": 288.14,
+      "grad_norm": 22.03531837463379,
+      "learning_rate": 3.953953953953954e-07,
+      "loss": 0.3025,
+      "step": 95950
+    },
+    {
+      "epoch": 288.17,
+      "grad_norm": 22.699960708618164,
+      "learning_rate": 3.943943943943944e-07,
+      "loss": 0.2578,
+      "step": 95960
+    },
+    {
+      "epoch": 288.2,
+      "grad_norm": 14.34776496887207,
+      "learning_rate": 3.9339339339339347e-07,
+      "loss": 0.3411,
+      "step": 95970
+    },
+    {
+      "epoch": 288.23,
+      "grad_norm": 13.548453330993652,
+      "learning_rate": 3.9239239239239244e-07,
+      "loss": 0.2752,
+      "step": 95980
+    },
+    {
+      "epoch": 288.26,
+      "grad_norm": 15.268775939941406,
+      "learning_rate": 3.9139139139139145e-07,
+      "loss": 0.3288,
+      "step": 95990
+    },
+    {
+      "epoch": 288.29,
+      "grad_norm": 18.63167381286621,
+      "learning_rate": 3.903903903903904e-07,
+      "loss": 0.3147,
+      "step": 96000
+    },
+    {
+      "epoch": 288.32,
+      "grad_norm": 9.698452949523926,
+      "learning_rate": 3.8938938938938944e-07,
+      "loss": 0.3151,
+      "step": 96010
+    },
+    {
+      "epoch": 288.35,
+      "grad_norm": 20.01169776916504,
+      "learning_rate": 3.883883883883884e-07,
+      "loss": 0.4085,
+      "step": 96020
+    },
+    {
+      "epoch": 288.38,
+      "grad_norm": 15.28549575805664,
+      "learning_rate": 3.873873873873874e-07,
+      "loss": 0.3749,
+      "step": 96030
+    },
+    {
+      "epoch": 288.41,
+      "grad_norm": 20.687519073486328,
+      "learning_rate": 3.863863863863864e-07,
+      "loss": 0.2905,
+      "step": 96040
+    },
+    {
+      "epoch": 288.44,
+      "grad_norm": 16.469724655151367,
+      "learning_rate": 3.853853853853854e-07,
+      "loss": 0.3358,
+      "step": 96050
+    },
+    {
+      "epoch": 288.47,
+      "grad_norm": 15.839497566223145,
+      "learning_rate": 3.843843843843844e-07,
+      "loss": 0.2836,
+      "step": 96060
+    },
+    {
+      "epoch": 288.5,
+      "grad_norm": 11.1560697555542,
+      "learning_rate": 3.833833833833834e-07,
+      "loss": 0.2605,
+      "step": 96070
+    },
+    {
+      "epoch": 288.53,
+      "grad_norm": 12.851758003234863,
+      "learning_rate": 3.823823823823824e-07,
+      "loss": 0.306,
+      "step": 96080
+    },
+    {
+      "epoch": 288.56,
+      "grad_norm": 11.820643424987793,
+      "learning_rate": 3.8138138138138137e-07,
+      "loss": 0.3135,
+      "step": 96090
+    },
+    {
+      "epoch": 288.59,
+      "grad_norm": 13.622471809387207,
+      "learning_rate": 3.8038038038038044e-07,
+      "loss": 0.3439,
+      "step": 96100
+    },
+    {
+      "epoch": 288.62,
+      "grad_norm": 13.533079147338867,
+      "learning_rate": 3.793793793793794e-07,
+      "loss": 0.3126,
+      "step": 96110
+    },
+    {
+      "epoch": 288.65,
+      "grad_norm": 21.12013816833496,
+      "learning_rate": 3.7837837837837843e-07,
+      "loss": 0.3623,
+      "step": 96120
+    },
+    {
+      "epoch": 288.68,
+      "grad_norm": 19.39162826538086,
+      "learning_rate": 3.7737737737737745e-07,
+      "loss": 0.3223,
+      "step": 96130
+    },
+    {
+      "epoch": 288.71,
+      "grad_norm": 17.669979095458984,
+      "learning_rate": 3.763763763763764e-07,
+      "loss": 0.2911,
+      "step": 96140
+    },
+    {
+      "epoch": 288.74,
+      "grad_norm": 21.441696166992188,
+      "learning_rate": 3.7537537537537543e-07,
+      "loss": 0.3088,
+      "step": 96150
+    },
+    {
+      "epoch": 288.77,
+      "grad_norm": 18.45435905456543,
+      "learning_rate": 3.743743743743744e-07,
+      "loss": 0.316,
+      "step": 96160
+    },
+    {
+      "epoch": 288.8,
+      "grad_norm": 18.043758392333984,
+      "learning_rate": 3.733733733733734e-07,
+      "loss": 0.3205,
+      "step": 96170
+    },
+    {
+      "epoch": 288.83,
+      "grad_norm": 23.376035690307617,
+      "learning_rate": 3.723723723723724e-07,
+      "loss": 0.2905,
+      "step": 96180
+    },
+    {
+      "epoch": 288.86,
+      "grad_norm": 22.983076095581055,
+      "learning_rate": 3.713713713713714e-07,
+      "loss": 0.3025,
+      "step": 96190
+    },
+    {
+      "epoch": 288.89,
+      "grad_norm": 22.729129791259766,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.3066,
+      "step": 96200
+    },
+    {
+      "epoch": 288.92,
+      "grad_norm": 12.635058403015137,
+      "learning_rate": 3.693693693693694e-07,
+      "loss": 0.3734,
+      "step": 96210
+    },
+    {
+      "epoch": 288.95,
+      "grad_norm": 16.63631248474121,
+      "learning_rate": 3.6836836836836835e-07,
+      "loss": 0.314,
+      "step": 96220
+    },
+    {
+      "epoch": 288.98,
+      "grad_norm": 19.34419822692871,
+      "learning_rate": 3.673673673673674e-07,
+      "loss": 0.3232,
+      "step": 96230
+    },
+    {
+      "epoch": 289.0,
+      "eval_accuracy": 0.8642,
+      "eval_loss": 0.5503892302513123,
+      "eval_runtime": 13.0004,
+      "eval_samples_per_second": 769.208,
+      "eval_steps_per_second": 3.077,
+      "step": 96237
+    },
+    {
+      "epoch": 289.01,
+      "grad_norm": 9.040816307067871,
+      "learning_rate": 3.6636636636636644e-07,
+      "loss": 0.278,
+      "step": 96240
+    },
+    {
+      "epoch": 289.04,
+      "grad_norm": 20.798120498657227,
+      "learning_rate": 3.653653653653654e-07,
+      "loss": 0.3591,
+      "step": 96250
+    },
+    {
+      "epoch": 289.07,
+      "grad_norm": 21.437618255615234,
+      "learning_rate": 3.643643643643644e-07,
+      "loss": 0.335,
+      "step": 96260
+    },
+    {
+      "epoch": 289.1,
+      "grad_norm": 17.988380432128906,
+      "learning_rate": 3.633633633633634e-07,
+      "loss": 0.2979,
+      "step": 96270
+    },
+    {
+      "epoch": 289.13,
+      "grad_norm": 19.805021286010742,
+      "learning_rate": 3.623623623623624e-07,
+      "loss": 0.2958,
+      "step": 96280
+    },
+    {
+      "epoch": 289.16,
+      "grad_norm": 16.728988647460938,
+      "learning_rate": 3.6136136136136137e-07,
+      "loss": 0.3226,
+      "step": 96290
+    },
+    {
+      "epoch": 289.19,
+      "grad_norm": 13.804011344909668,
+      "learning_rate": 3.603603603603604e-07,
+      "loss": 0.2982,
+      "step": 96300
+    },
+    {
+      "epoch": 289.22,
+      "grad_norm": 11.935112953186035,
+      "learning_rate": 3.593593593593594e-07,
+      "loss": 0.3291,
+      "step": 96310
+    },
+    {
+      "epoch": 289.25,
+      "grad_norm": 13.47002124786377,
+      "learning_rate": 3.5835835835835837e-07,
+      "loss": 0.3603,
+      "step": 96320
+    },
+    {
+      "epoch": 289.28,
+      "grad_norm": 20.98748779296875,
+      "learning_rate": 3.573573573573574e-07,
+      "loss": 0.3365,
+      "step": 96330
+    },
+    {
+      "epoch": 289.31,
+      "grad_norm": 13.285073280334473,
+      "learning_rate": 3.5635635635635636e-07,
+      "loss": 0.2926,
+      "step": 96340
+    },
+    {
+      "epoch": 289.34,
+      "grad_norm": 14.893031120300293,
+      "learning_rate": 3.553553553553554e-07,
+      "loss": 0.3669,
+      "step": 96350
+    },
+    {
+      "epoch": 289.37,
+      "grad_norm": 21.61373519897461,
+      "learning_rate": 3.5435435435435434e-07,
+      "loss": 0.291,
+      "step": 96360
+    },
+    {
+      "epoch": 289.4,
+      "grad_norm": 17.75531005859375,
+      "learning_rate": 3.533533533533534e-07,
+      "loss": 0.3272,
+      "step": 96370
+    },
+    {
+      "epoch": 289.43,
+      "grad_norm": 23.97504234313965,
+      "learning_rate": 3.5235235235235243e-07,
+      "loss": 0.2824,
+      "step": 96380
+    },
+    {
+      "epoch": 289.46,
+      "grad_norm": 13.625621795654297,
+      "learning_rate": 3.513513513513514e-07,
+      "loss": 0.3053,
+      "step": 96390
+    },
+    {
+      "epoch": 289.49,
+      "grad_norm": 13.818299293518066,
+      "learning_rate": 3.503503503503504e-07,
+      "loss": 0.3673,
+      "step": 96400
+    },
+    {
+      "epoch": 289.52,
+      "grad_norm": 17.596343994140625,
+      "learning_rate": 3.493493493493494e-07,
+      "loss": 0.2995,
+      "step": 96410
+    },
+    {
+      "epoch": 289.55,
+      "grad_norm": 13.782303810119629,
+      "learning_rate": 3.483483483483484e-07,
+      "loss": 0.3249,
+      "step": 96420
+    },
+    {
+      "epoch": 289.58,
+      "grad_norm": 15.764554023742676,
+      "learning_rate": 3.4734734734734736e-07,
+      "loss": 0.3243,
+      "step": 96430
+    },
+    {
+      "epoch": 289.61,
+      "grad_norm": 19.799156188964844,
+      "learning_rate": 3.463463463463464e-07,
+      "loss": 0.2886,
+      "step": 96440
+    },
+    {
+      "epoch": 289.64,
+      "grad_norm": 11.604363441467285,
+      "learning_rate": 3.4534534534534535e-07,
+      "loss": 0.2948,
+      "step": 96450
+    },
+    {
+      "epoch": 289.67,
+      "grad_norm": 13.688623428344727,
+      "learning_rate": 3.4434434434434436e-07,
+      "loss": 0.3274,
+      "step": 96460
+    },
+    {
+      "epoch": 289.7,
+      "grad_norm": 16.79641342163086,
+      "learning_rate": 3.4334334334334333e-07,
+      "loss": 0.3076,
+      "step": 96470
+    },
+    {
+      "epoch": 289.73,
+      "grad_norm": 22.610092163085938,
+      "learning_rate": 3.4234234234234235e-07,
+      "loss": 0.34,
+      "step": 96480
+    },
+    {
+      "epoch": 289.76,
+      "grad_norm": 12.2433500289917,
+      "learning_rate": 3.413413413413413e-07,
+      "loss": 0.2966,
+      "step": 96490
+    },
+    {
+      "epoch": 289.79,
+      "grad_norm": 16.55497169494629,
+      "learning_rate": 3.403403403403404e-07,
+      "loss": 0.3567,
+      "step": 96500
+    },
+    {
+      "epoch": 289.82,
+      "grad_norm": 14.230645179748535,
+      "learning_rate": 3.393393393393394e-07,
+      "loss": 0.3092,
+      "step": 96510
+    },
+    {
+      "epoch": 289.85,
+      "grad_norm": 15.637765884399414,
+      "learning_rate": 3.3833833833833837e-07,
+      "loss": 0.3244,
+      "step": 96520
+    },
+    {
+      "epoch": 289.88,
+      "grad_norm": 12.426387786865234,
+      "learning_rate": 3.373373373373374e-07,
+      "loss": 0.3677,
+      "step": 96530
+    },
+    {
+      "epoch": 289.91,
+      "grad_norm": 18.772769927978516,
+      "learning_rate": 3.3633633633633635e-07,
+      "loss": 0.3097,
+      "step": 96540
+    },
+    {
+      "epoch": 289.94,
+      "grad_norm": 13.57267951965332,
+      "learning_rate": 3.3533533533533537e-07,
+      "loss": 0.3,
+      "step": 96550
+    },
+    {
+      "epoch": 289.97,
+      "grad_norm": 19.251813888549805,
+      "learning_rate": 3.343343343343344e-07,
+      "loss": 0.3178,
+      "step": 96560
+    },
+    {
+      "epoch": 290.0,
+      "grad_norm": 0.1787324994802475,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 0.286,
+      "step": 96570
+    },
+    {
+      "epoch": 290.0,
+      "eval_accuracy": 0.8644,
+      "eval_loss": 0.5499082803726196,
+      "eval_runtime": 13.2127,
+      "eval_samples_per_second": 756.849,
+      "eval_steps_per_second": 3.027,
+      "step": 96570
+    },
+    {
+      "epoch": 290.03,
+      "grad_norm": 15.329418182373047,
+      "learning_rate": 3.3233233233233237e-07,
+      "loss": 0.3103,
+      "step": 96580
+    },
+    {
+      "epoch": 290.06,
+      "grad_norm": 13.284131050109863,
+      "learning_rate": 3.3133133133133134e-07,
+      "loss": 0.3265,
+      "step": 96590
+    },
+    {
+      "epoch": 290.09,
+      "grad_norm": 16.927453994750977,
+      "learning_rate": 3.3033033033033036e-07,
+      "loss": 0.3063,
+      "step": 96600
+    },
+    {
+      "epoch": 290.12,
+      "grad_norm": 20.412948608398438,
+      "learning_rate": 3.293293293293293e-07,
+      "loss": 0.3398,
+      "step": 96610
+    },
+    {
+      "epoch": 290.15,
+      "grad_norm": 13.6939697265625,
+      "learning_rate": 3.2832832832832834e-07,
+      "loss": 0.3625,
+      "step": 96620
+    },
+    {
+      "epoch": 290.18,
+      "grad_norm": 12.929060935974121,
+      "learning_rate": 3.273273273273274e-07,
+      "loss": 0.285,
+      "step": 96630
+    },
+    {
+      "epoch": 290.21,
+      "grad_norm": 21.821401596069336,
+      "learning_rate": 3.263263263263264e-07,
+      "loss": 0.3322,
+      "step": 96640
+    },
+    {
+      "epoch": 290.24,
+      "grad_norm": 13.27804946899414,
+      "learning_rate": 3.253253253253254e-07,
+      "loss": 0.3244,
+      "step": 96650
+    },
+    {
+      "epoch": 290.27,
+      "grad_norm": 19.736305236816406,
+      "learning_rate": 3.2432432432432436e-07,
+      "loss": 0.3113,
+      "step": 96660
+    },
+    {
+      "epoch": 290.3,
+      "grad_norm": 14.0823974609375,
+      "learning_rate": 3.233233233233234e-07,
+      "loss": 0.3641,
+      "step": 96670
+    },
+    {
+      "epoch": 290.33,
+      "grad_norm": 15.551570892333984,
+      "learning_rate": 3.2232232232232234e-07,
+      "loss": 0.3233,
+      "step": 96680
+    },
+    {
+      "epoch": 290.36,
+      "grad_norm": 13.124831199645996,
+      "learning_rate": 3.2132132132132136e-07,
+      "loss": 0.3125,
+      "step": 96690
+    },
+    {
+      "epoch": 290.39,
+      "grad_norm": 17.88973617553711,
+      "learning_rate": 3.2032032032032033e-07,
+      "loss": 0.3526,
+      "step": 96700
+    },
+    {
+      "epoch": 290.42,
+      "grad_norm": 16.066194534301758,
+      "learning_rate": 3.1931931931931935e-07,
+      "loss": 0.3288,
+      "step": 96710
+    },
+    {
+      "epoch": 290.45,
+      "grad_norm": 21.549394607543945,
+      "learning_rate": 3.183183183183183e-07,
+      "loss": 0.3524,
+      "step": 96720
+    },
+    {
+      "epoch": 290.48,
+      "grad_norm": 12.73495864868164,
+      "learning_rate": 3.1731731731731733e-07,
+      "loss": 0.3393,
+      "step": 96730
+    },
+    {
+      "epoch": 290.51,
+      "grad_norm": 14.1395263671875,
+      "learning_rate": 3.163163163163163e-07,
+      "loss": 0.3275,
+      "step": 96740
+    },
+    {
+      "epoch": 290.54,
+      "grad_norm": 18.338144302368164,
+      "learning_rate": 3.153153153153153e-07,
+      "loss": 0.3463,
+      "step": 96750
+    },
+    {
+      "epoch": 290.57,
+      "grad_norm": 24.38106346130371,
+      "learning_rate": 3.1431431431431433e-07,
+      "loss": 0.2798,
+      "step": 96760
+    },
+    {
+      "epoch": 290.6,
+      "grad_norm": 12.330666542053223,
+      "learning_rate": 3.1331331331331335e-07,
+      "loss": 0.3054,
+      "step": 96770
+    },
+    {
+      "epoch": 290.63,
+      "grad_norm": 16.583213806152344,
+      "learning_rate": 3.123123123123123e-07,
+      "loss": 0.3545,
+      "step": 96780
+    },
+    {
+      "epoch": 290.66,
+      "grad_norm": 14.47153091430664,
+      "learning_rate": 3.1131131131131133e-07,
+      "loss": 0.2979,
+      "step": 96790
+    },
+    {
+      "epoch": 290.69,
+      "grad_norm": 18.447267532348633,
+      "learning_rate": 3.1031031031031035e-07,
+      "loss": 0.3188,
+      "step": 96800
+    },
+    {
+      "epoch": 290.72,
+      "grad_norm": 11.022024154663086,
+      "learning_rate": 3.0930930930930937e-07,
+      "loss": 0.3171,
+      "step": 96810
+    },
+    {
+      "epoch": 290.75,
+      "grad_norm": 18.270069122314453,
+      "learning_rate": 3.0830830830830834e-07,
+      "loss": 0.3218,
+      "step": 96820
+    },
+    {
+      "epoch": 290.78,
+      "grad_norm": 14.88742446899414,
+      "learning_rate": 3.0730730730730735e-07,
+      "loss": 0.3206,
+      "step": 96830
+    },
+    {
+      "epoch": 290.81,
+      "grad_norm": 12.474881172180176,
+      "learning_rate": 3.063063063063063e-07,
+      "loss": 0.3575,
+      "step": 96840
+    },
+    {
+      "epoch": 290.84,
+      "grad_norm": 14.712125778198242,
+      "learning_rate": 3.0530530530530534e-07,
+      "loss": 0.3232,
+      "step": 96850
+    },
+    {
+      "epoch": 290.87,
+      "grad_norm": 17.7692813873291,
+      "learning_rate": 3.043043043043043e-07,
+      "loss": 0.3836,
+      "step": 96860
+    },
+    {
+      "epoch": 290.9,
+      "grad_norm": 15.043848037719727,
+      "learning_rate": 3.033033033033033e-07,
+      "loss": 0.304,
+      "step": 96870
+    },
+    {
+      "epoch": 290.93,
+      "grad_norm": 16.228734970092773,
+      "learning_rate": 3.0230230230230234e-07,
+      "loss": 0.3426,
+      "step": 96880
+    },
+    {
+      "epoch": 290.96,
+      "grad_norm": 12.628997802734375,
+      "learning_rate": 3.013013013013013e-07,
+      "loss": 0.3076,
+      "step": 96890
+    },
+    {
+      "epoch": 290.99,
+      "grad_norm": 14.354206085205078,
+      "learning_rate": 3.003003003003003e-07,
+      "loss": 0.2953,
+      "step": 96900
+    },
+    {
+      "epoch": 291.0,
+      "eval_accuracy": 0.8638,
+      "eval_loss": 0.5503947138786316,
+      "eval_runtime": 13.0293,
+      "eval_samples_per_second": 767.499,
+      "eval_steps_per_second": 3.07,
+      "step": 96903
+    },
+    {
+      "epoch": 291.02,
+      "grad_norm": 13.490812301635742,
+      "learning_rate": 2.9929929929929934e-07,
+      "loss": 0.3905,
+      "step": 96910
+    },
+    {
+      "epoch": 291.05,
+      "grad_norm": 15.23181438446045,
+      "learning_rate": 2.982982982982983e-07,
+      "loss": 0.3213,
+      "step": 96920
+    },
+    {
+      "epoch": 291.08,
+      "grad_norm": 19.434141159057617,
+      "learning_rate": 2.972972972972973e-07,
+      "loss": 0.3143,
+      "step": 96930
+    },
+    {
+      "epoch": 291.11,
+      "grad_norm": 17.34665298461914,
+      "learning_rate": 2.9629629629629634e-07,
+      "loss": 0.3261,
+      "step": 96940
+    },
+    {
+      "epoch": 291.14,
+      "grad_norm": 15.821691513061523,
+      "learning_rate": 2.952952952952953e-07,
+      "loss": 0.2608,
+      "step": 96950
+    },
+    {
+      "epoch": 291.17,
+      "grad_norm": 14.661909103393555,
+      "learning_rate": 2.9429429429429433e-07,
+      "loss": 0.3461,
+      "step": 96960
+    },
+    {
+      "epoch": 291.2,
+      "grad_norm": 23.258848190307617,
+      "learning_rate": 2.932932932932933e-07,
+      "loss": 0.2714,
+      "step": 96970
+    },
+    {
+      "epoch": 291.23,
+      "grad_norm": 17.390119552612305,
+      "learning_rate": 2.922922922922923e-07,
+      "loss": 0.3001,
+      "step": 96980
+    },
+    {
+      "epoch": 291.26,
+      "grad_norm": 17.6117000579834,
+      "learning_rate": 2.912912912912913e-07,
+      "loss": 0.3099,
+      "step": 96990
+    },
+    {
+      "epoch": 291.29,
+      "grad_norm": 17.13506317138672,
+      "learning_rate": 2.9029029029029035e-07,
+      "loss": 0.3249,
+      "step": 97000
+    },
+    {
+      "epoch": 291.32,
+      "grad_norm": 13.631427764892578,
+      "learning_rate": 2.892892892892893e-07,
+      "loss": 0.3141,
+      "step": 97010
+    },
+    {
+      "epoch": 291.35,
+      "grad_norm": 23.62021255493164,
+      "learning_rate": 2.8828828828828833e-07,
+      "loss": 0.3759,
+      "step": 97020
+    },
+    {
+      "epoch": 291.38,
+      "grad_norm": 15.935111045837402,
+      "learning_rate": 2.872872872872873e-07,
+      "loss": 0.342,
+      "step": 97030
+    },
+    {
+      "epoch": 291.41,
+      "grad_norm": 17.429492950439453,
+      "learning_rate": 2.862862862862863e-07,
+      "loss": 0.3427,
+      "step": 97040
+    },
+    {
+      "epoch": 291.44,
+      "grad_norm": 14.91895866394043,
+      "learning_rate": 2.852852852852853e-07,
+      "loss": 0.3651,
+      "step": 97050
+    },
+    {
+      "epoch": 291.47,
+      "grad_norm": 22.260498046875,
+      "learning_rate": 2.842842842842843e-07,
+      "loss": 0.2901,
+      "step": 97060
+    },
+    {
+      "epoch": 291.5,
+      "grad_norm": 13.855570793151855,
+      "learning_rate": 2.832832832832833e-07,
+      "loss": 0.2473,
+      "step": 97070
+    },
+    {
+      "epoch": 291.53,
+      "grad_norm": 13.888151168823242,
+      "learning_rate": 2.8228228228228234e-07,
+      "loss": 0.4189,
+      "step": 97080
+    },
+    {
+      "epoch": 291.56,
+      "grad_norm": 7.7561564445495605,
+      "learning_rate": 2.812812812812813e-07,
+      "loss": 0.3282,
+      "step": 97090
+    },
+    {
+      "epoch": 291.59,
+      "grad_norm": 14.18308162689209,
+      "learning_rate": 2.802802802802803e-07,
+      "loss": 0.2976,
+      "step": 97100
+    },
+    {
+      "epoch": 291.62,
+      "grad_norm": 21.238554000854492,
+      "learning_rate": 2.792792792792793e-07,
+      "loss": 0.3154,
+      "step": 97110
+    },
+    {
+      "epoch": 291.65,
+      "grad_norm": 17.759687423706055,
+      "learning_rate": 2.782782782782783e-07,
+      "loss": 0.3377,
+      "step": 97120
+    },
+    {
+      "epoch": 291.68,
+      "grad_norm": 10.516297340393066,
+      "learning_rate": 2.772772772772773e-07,
+      "loss": 0.2765,
+      "step": 97130
+    },
+    {
+      "epoch": 291.71,
+      "grad_norm": 15.048307418823242,
+      "learning_rate": 2.762762762762763e-07,
+      "loss": 0.2867,
+      "step": 97140
+    },
+    {
+      "epoch": 291.74,
+      "grad_norm": 16.53089714050293,
+      "learning_rate": 2.752752752752753e-07,
+      "loss": 0.3274,
+      "step": 97150
+    },
+    {
+      "epoch": 291.77,
+      "grad_norm": 19.13397979736328,
+      "learning_rate": 2.742742742742743e-07,
+      "loss": 0.2894,
+      "step": 97160
+    },
+    {
+      "epoch": 291.8,
+      "grad_norm": 15.786242485046387,
+      "learning_rate": 2.732732732732733e-07,
+      "loss": 0.3092,
+      "step": 97170
+    },
+    {
+      "epoch": 291.83,
+      "grad_norm": 9.784735679626465,
+      "learning_rate": 2.722722722722723e-07,
+      "loss": 0.3199,
+      "step": 97180
+    },
+    {
+      "epoch": 291.86,
+      "grad_norm": 18.6018009185791,
+      "learning_rate": 2.7127127127127127e-07,
+      "loss": 0.3377,
+      "step": 97190
+    },
+    {
+      "epoch": 291.89,
+      "grad_norm": 18.438369750976562,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.3343,
+      "step": 97200
+    },
+    {
+      "epoch": 291.92,
+      "grad_norm": 18.94111442565918,
+      "learning_rate": 2.692692692692693e-07,
+      "loss": 0.3197,
+      "step": 97210
+    },
+    {
+      "epoch": 291.95,
+      "grad_norm": 17.095130920410156,
+      "learning_rate": 2.682682682682683e-07,
+      "loss": 0.2935,
+      "step": 97220
+    },
+    {
+      "epoch": 291.98,
+      "grad_norm": 12.384288787841797,
+      "learning_rate": 2.672672672672673e-07,
+      "loss": 0.3371,
+      "step": 97230
+    },
+    {
+      "epoch": 292.0,
+      "eval_accuracy": 0.8645,
+      "eval_loss": 0.5511572957038879,
+      "eval_runtime": 12.8477,
+      "eval_samples_per_second": 778.347,
+      "eval_steps_per_second": 3.113,
+      "step": 97236
+    },
+    {
+      "epoch": 292.01,
+      "grad_norm": 13.935205459594727,
+      "learning_rate": 2.6626626626626626e-07,
+      "loss": 0.3079,
+      "step": 97240
+    },
+    {
+      "epoch": 292.04,
+      "grad_norm": 25.830204010009766,
+      "learning_rate": 2.652652652652653e-07,
+      "loss": 0.3141,
+      "step": 97250
+    },
+    {
+      "epoch": 292.07,
+      "grad_norm": 17.413372039794922,
+      "learning_rate": 2.642642642642643e-07,
+      "loss": 0.4115,
+      "step": 97260
+    },
+    {
+      "epoch": 292.1,
+      "grad_norm": 18.23711585998535,
+      "learning_rate": 2.632632632632633e-07,
+      "loss": 0.3034,
+      "step": 97270
+    },
+    {
+      "epoch": 292.13,
+      "grad_norm": 14.453680992126465,
+      "learning_rate": 2.622622622622623e-07,
+      "loss": 0.3589,
+      "step": 97280
+    },
+    {
+      "epoch": 292.16,
+      "grad_norm": 15.717905044555664,
+      "learning_rate": 2.612612612612613e-07,
+      "loss": 0.3181,
+      "step": 97290
+    },
+    {
+      "epoch": 292.19,
+      "grad_norm": 17.636072158813477,
+      "learning_rate": 2.6026026026026026e-07,
+      "loss": 0.3234,
+      "step": 97300
+    },
+    {
+      "epoch": 292.22,
+      "grad_norm": 16.873214721679688,
+      "learning_rate": 2.592592592592593e-07,
+      "loss": 0.3993,
+      "step": 97310
+    },
+    {
+      "epoch": 292.25,
+      "grad_norm": 24.99449920654297,
+      "learning_rate": 2.5825825825825825e-07,
+      "loss": 0.3186,
+      "step": 97320
+    },
+    {
+      "epoch": 292.28,
+      "grad_norm": 18.016176223754883,
+      "learning_rate": 2.572572572572573e-07,
+      "loss": 0.2974,
+      "step": 97330
+    },
+    {
+      "epoch": 292.31,
+      "grad_norm": 14.650148391723633,
+      "learning_rate": 2.562562562562563e-07,
+      "loss": 0.3726,
+      "step": 97340
+    },
+    {
+      "epoch": 292.34,
+      "grad_norm": 12.884119033813477,
+      "learning_rate": 2.552552552552553e-07,
+      "loss": 0.291,
+      "step": 97350
+    },
+    {
+      "epoch": 292.37,
+      "grad_norm": 22.29964828491211,
+      "learning_rate": 2.5425425425425427e-07,
+      "loss": 0.3283,
+      "step": 97360
+    },
+    {
+      "epoch": 292.4,
+      "grad_norm": 14.124082565307617,
+      "learning_rate": 2.532532532532533e-07,
+      "loss": 0.3271,
+      "step": 97370
+    },
+    {
+      "epoch": 292.43,
+      "grad_norm": 16.523113250732422,
+      "learning_rate": 2.5225225225225225e-07,
+      "loss": 0.2471,
+      "step": 97380
+    },
+    {
+      "epoch": 292.46,
+      "grad_norm": 12.488131523132324,
+      "learning_rate": 2.5125125125125127e-07,
+      "loss": 0.3369,
+      "step": 97390
+    },
+    {
+      "epoch": 292.49,
+      "grad_norm": 16.54545021057129,
+      "learning_rate": 2.502502502502503e-07,
+      "loss": 0.38,
+      "step": 97400
+    },
+    {
+      "epoch": 292.52,
+      "grad_norm": 15.178956985473633,
+      "learning_rate": 2.492492492492493e-07,
+      "loss": 0.2832,
+      "step": 97410
+    },
+    {
+      "epoch": 292.55,
+      "grad_norm": 15.353004455566406,
+      "learning_rate": 2.4824824824824827e-07,
+      "loss": 0.3872,
+      "step": 97420
+    },
+    {
+      "epoch": 292.58,
+      "grad_norm": 14.343247413635254,
+      "learning_rate": 2.472472472472473e-07,
+      "loss": 0.2989,
+      "step": 97430
+    },
+    {
+      "epoch": 292.61,
+      "grad_norm": 19.888219833374023,
+      "learning_rate": 2.4624624624624625e-07,
+      "loss": 0.2861,
+      "step": 97440
+    },
+    {
+      "epoch": 292.64,
+      "grad_norm": 12.43703556060791,
+      "learning_rate": 2.4524524524524527e-07,
+      "loss": 0.2917,
+      "step": 97450
+    },
+    {
+      "epoch": 292.67,
+      "grad_norm": 21.266767501831055,
+      "learning_rate": 2.4424424424424424e-07,
+      "loss": 0.3194,
+      "step": 97460
+    },
+    {
+      "epoch": 292.7,
+      "grad_norm": 15.503408432006836,
+      "learning_rate": 2.4324324324324326e-07,
+      "loss": 0.3115,
+      "step": 97470
+    },
+    {
+      "epoch": 292.73,
+      "grad_norm": 17.56169319152832,
+      "learning_rate": 2.422422422422423e-07,
+      "loss": 0.3063,
+      "step": 97480
+    },
+    {
+      "epoch": 292.76,
+      "grad_norm": 11.173048973083496,
+      "learning_rate": 2.4124124124124124e-07,
+      "loss": 0.3254,
+      "step": 97490
+    },
+    {
+      "epoch": 292.79,
+      "grad_norm": 11.170706748962402,
+      "learning_rate": 2.4024024024024026e-07,
+      "loss": 0.2815,
+      "step": 97500
+    },
+    {
+      "epoch": 292.82,
+      "grad_norm": 17.72882843017578,
+      "learning_rate": 2.392392392392393e-07,
+      "loss": 0.3302,
+      "step": 97510
+    },
+    {
+      "epoch": 292.85,
+      "grad_norm": 17.764375686645508,
+      "learning_rate": 2.3823823823823824e-07,
+      "loss": 0.2796,
+      "step": 97520
+    },
+    {
+      "epoch": 292.88,
+      "grad_norm": 17.335556030273438,
+      "learning_rate": 2.3723723723723726e-07,
+      "loss": 0.3171,
+      "step": 97530
+    },
+    {
+      "epoch": 292.91,
+      "grad_norm": 17.362009048461914,
+      "learning_rate": 2.3623623623623628e-07,
+      "loss": 0.298,
+      "step": 97540
+    },
+    {
+      "epoch": 292.94,
+      "grad_norm": 20.488506317138672,
+      "learning_rate": 2.3523523523523527e-07,
+      "loss": 0.3305,
+      "step": 97550
+    },
+    {
+      "epoch": 292.97,
+      "grad_norm": 14.521987915039062,
+      "learning_rate": 2.3423423423423426e-07,
+      "loss": 0.3631,
+      "step": 97560
+    },
+    {
+      "epoch": 293.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.5504023432731628,
+      "eval_runtime": 13.1081,
+      "eval_samples_per_second": 762.889,
+      "eval_steps_per_second": 3.052,
+      "step": 97569
+    },
+    {
+      "epoch": 293.0,
+      "grad_norm": 18.95486831665039,
+      "learning_rate": 2.3323323323323325e-07,
+      "loss": 0.3174,
+      "step": 97570
+    },
+    {
+      "epoch": 293.03,
+      "grad_norm": 13.639264106750488,
+      "learning_rate": 2.3223223223223225e-07,
+      "loss": 0.2945,
+      "step": 97580
+    },
+    {
+      "epoch": 293.06,
+      "grad_norm": 11.522398948669434,
+      "learning_rate": 2.3123123123123124e-07,
+      "loss": 0.2877,
+      "step": 97590
+    },
+    {
+      "epoch": 293.09,
+      "grad_norm": 17.328575134277344,
+      "learning_rate": 2.3023023023023026e-07,
+      "loss": 0.2985,
+      "step": 97600
+    },
+    {
+      "epoch": 293.12,
+      "grad_norm": 12.14549446105957,
+      "learning_rate": 2.2922922922922925e-07,
+      "loss": 0.2823,
+      "step": 97610
+    },
+    {
+      "epoch": 293.15,
+      "grad_norm": 12.818968772888184,
+      "learning_rate": 2.2822822822822824e-07,
+      "loss": 0.3184,
+      "step": 97620
+    },
+    {
+      "epoch": 293.18,
+      "grad_norm": 20.425270080566406,
+      "learning_rate": 2.2722722722722726e-07,
+      "loss": 0.3453,
+      "step": 97630
+    },
+    {
+      "epoch": 293.21,
+      "grad_norm": 9.232460975646973,
+      "learning_rate": 2.2622622622622625e-07,
+      "loss": 0.3089,
+      "step": 97640
+    },
+    {
+      "epoch": 293.24,
+      "grad_norm": 10.256820678710938,
+      "learning_rate": 2.2522522522522524e-07,
+      "loss": 0.3328,
+      "step": 97650
+    },
+    {
+      "epoch": 293.27,
+      "grad_norm": 14.281722068786621,
+      "learning_rate": 2.2422422422422423e-07,
+      "loss": 0.3336,
+      "step": 97660
+    },
+    {
+      "epoch": 293.3,
+      "grad_norm": 18.69748306274414,
+      "learning_rate": 2.2322322322322325e-07,
+      "loss": 0.3098,
+      "step": 97670
+    },
+    {
+      "epoch": 293.33,
+      "grad_norm": 14.379137992858887,
+      "learning_rate": 2.2222222222222224e-07,
+      "loss": 0.3455,
+      "step": 97680
+    },
+    {
+      "epoch": 293.36,
+      "grad_norm": 11.945849418640137,
+      "learning_rate": 2.2122122122122124e-07,
+      "loss": 0.3198,
+      "step": 97690
+    },
+    {
+      "epoch": 293.39,
+      "grad_norm": 14.345710754394531,
+      "learning_rate": 2.2022022022022023e-07,
+      "loss": 0.3562,
+      "step": 97700
+    },
+    {
+      "epoch": 293.42,
+      "grad_norm": 13.175491333007812,
+      "learning_rate": 2.1921921921921922e-07,
+      "loss": 0.3135,
+      "step": 97710
+    },
+    {
+      "epoch": 293.45,
+      "grad_norm": 11.6928129196167,
+      "learning_rate": 2.182182182182182e-07,
+      "loss": 0.3419,
+      "step": 97720
+    },
+    {
+      "epoch": 293.48,
+      "grad_norm": 16.029081344604492,
+      "learning_rate": 2.1721721721721726e-07,
+      "loss": 0.3539,
+      "step": 97730
+    },
+    {
+      "epoch": 293.51,
+      "grad_norm": 16.71297264099121,
+      "learning_rate": 2.1621621621621625e-07,
+      "loss": 0.3305,
+      "step": 97740
+    },
+    {
+      "epoch": 293.54,
+      "grad_norm": 18.25401496887207,
+      "learning_rate": 2.1521521521521524e-07,
+      "loss": 0.332,
+      "step": 97750
+    },
+    {
+      "epoch": 293.57,
+      "grad_norm": 15.816332817077637,
+      "learning_rate": 2.1421421421421423e-07,
+      "loss": 0.2852,
+      "step": 97760
+    },
+    {
+      "epoch": 293.6,
+      "grad_norm": 12.033167839050293,
+      "learning_rate": 2.1321321321321322e-07,
+      "loss": 0.277,
+      "step": 97770
+    },
+    {
+      "epoch": 293.63,
+      "grad_norm": 16.083524703979492,
+      "learning_rate": 2.1221221221221222e-07,
+      "loss": 0.3211,
+      "step": 97780
+    },
+    {
+      "epoch": 293.66,
+      "grad_norm": 15.669189453125,
+      "learning_rate": 2.112112112112112e-07,
+      "loss": 0.2891,
+      "step": 97790
+    },
+    {
+      "epoch": 293.69,
+      "grad_norm": 13.994531631469727,
+      "learning_rate": 2.1021021021021025e-07,
+      "loss": 0.32,
+      "step": 97800
+    },
+    {
+      "epoch": 293.72,
+      "grad_norm": 15.890456199645996,
+      "learning_rate": 2.0920920920920924e-07,
+      "loss": 0.3098,
+      "step": 97810
+    },
+    {
+      "epoch": 293.75,
+      "grad_norm": 16.147003173828125,
+      "learning_rate": 2.0820820820820824e-07,
+      "loss": 0.3392,
+      "step": 97820
+    },
+    {
+      "epoch": 293.78,
+      "grad_norm": 18.92282485961914,
+      "learning_rate": 2.0720720720720723e-07,
+      "loss": 0.3708,
+      "step": 97830
+    },
+    {
+      "epoch": 293.81,
+      "grad_norm": 20.30705451965332,
+      "learning_rate": 2.0620620620620622e-07,
+      "loss": 0.3012,
+      "step": 97840
+    },
+    {
+      "epoch": 293.84,
+      "grad_norm": 10.15685749053955,
+      "learning_rate": 2.052052052052052e-07,
+      "loss": 0.3378,
+      "step": 97850
+    },
+    {
+      "epoch": 293.87,
+      "grad_norm": 14.019715309143066,
+      "learning_rate": 2.042042042042042e-07,
+      "loss": 0.3064,
+      "step": 97860
+    },
+    {
+      "epoch": 293.9,
+      "grad_norm": 27.352989196777344,
+      "learning_rate": 2.0320320320320322e-07,
+      "loss": 0.3085,
+      "step": 97870
+    },
+    {
+      "epoch": 293.93,
+      "grad_norm": 15.127116203308105,
+      "learning_rate": 2.0220220220220224e-07,
+      "loss": 0.3096,
+      "step": 97880
+    },
+    {
+      "epoch": 293.96,
+      "grad_norm": 17.54540252685547,
+      "learning_rate": 2.0120120120120123e-07,
+      "loss": 0.2994,
+      "step": 97890
+    },
+    {
+      "epoch": 293.99,
+      "grad_norm": 15.755694389343262,
+      "learning_rate": 2.0020020020020022e-07,
+      "loss": 0.3439,
+      "step": 97900
+    },
+    {
+      "epoch": 294.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5495803356170654,
+      "eval_runtime": 12.6907,
+      "eval_samples_per_second": 787.98,
+      "eval_steps_per_second": 3.152,
+      "step": 97902
+    },
+    {
+      "epoch": 294.02,
+      "grad_norm": 16.544973373413086,
+      "learning_rate": 1.9919919919919922e-07,
+      "loss": 0.2457,
+      "step": 97910
+    },
+    {
+      "epoch": 294.05,
+      "grad_norm": 13.580425262451172,
+      "learning_rate": 1.981981981981982e-07,
+      "loss": 0.309,
+      "step": 97920
+    },
+    {
+      "epoch": 294.08,
+      "grad_norm": 16.530502319335938,
+      "learning_rate": 1.971971971971972e-07,
+      "loss": 0.3475,
+      "step": 97930
+    },
+    {
+      "epoch": 294.11,
+      "grad_norm": 19.475366592407227,
+      "learning_rate": 1.9619619619619622e-07,
+      "loss": 0.3077,
+      "step": 97940
+    },
+    {
+      "epoch": 294.14,
+      "grad_norm": 12.679266929626465,
+      "learning_rate": 1.951951951951952e-07,
+      "loss": 0.35,
+      "step": 97950
+    },
+    {
+      "epoch": 294.17,
+      "grad_norm": 12.343378067016602,
+      "learning_rate": 1.941941941941942e-07,
+      "loss": 0.3148,
+      "step": 97960
+    },
+    {
+      "epoch": 294.2,
+      "grad_norm": 11.124180793762207,
+      "learning_rate": 1.931931931931932e-07,
+      "loss": 0.3274,
+      "step": 97970
+    },
+    {
+      "epoch": 294.23,
+      "grad_norm": 20.661916732788086,
+      "learning_rate": 1.921921921921922e-07,
+      "loss": 0.3452,
+      "step": 97980
+    },
+    {
+      "epoch": 294.26,
+      "grad_norm": 22.290983200073242,
+      "learning_rate": 1.911911911911912e-07,
+      "loss": 0.3146,
+      "step": 97990
+    },
+    {
+      "epoch": 294.29,
+      "grad_norm": 8.223438262939453,
+      "learning_rate": 1.9019019019019022e-07,
+      "loss": 0.3599,
+      "step": 98000
+    },
+    {
+      "epoch": 294.32,
+      "grad_norm": 13.177309036254883,
+      "learning_rate": 1.8918918918918921e-07,
+      "loss": 0.2999,
+      "step": 98010
+    },
+    {
+      "epoch": 294.35,
+      "grad_norm": 16.822629928588867,
+      "learning_rate": 1.881881881881882e-07,
+      "loss": 0.2887,
+      "step": 98020
+    },
+    {
+      "epoch": 294.38,
+      "grad_norm": 16.006317138671875,
+      "learning_rate": 1.871871871871872e-07,
+      "loss": 0.3199,
+      "step": 98030
+    },
+    {
+      "epoch": 294.41,
+      "grad_norm": 12.193438529968262,
+      "learning_rate": 1.861861861861862e-07,
+      "loss": 0.3152,
+      "step": 98040
+    },
+    {
+      "epoch": 294.44,
+      "grad_norm": 12.712474822998047,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 0.3298,
+      "step": 98050
+    },
+    {
+      "epoch": 294.47,
+      "grad_norm": 16.264141082763672,
+      "learning_rate": 1.8418418418418417e-07,
+      "loss": 0.2941,
+      "step": 98060
+    },
+    {
+      "epoch": 294.5,
+      "grad_norm": 17.245784759521484,
+      "learning_rate": 1.8318318318318322e-07,
+      "loss": 0.3108,
+      "step": 98070
+    },
+    {
+      "epoch": 294.53,
+      "grad_norm": 12.014209747314453,
+      "learning_rate": 1.821821821821822e-07,
+      "loss": 0.3211,
+      "step": 98080
+    },
+    {
+      "epoch": 294.56,
+      "grad_norm": 17.100343704223633,
+      "learning_rate": 1.811811811811812e-07,
+      "loss": 0.3221,
+      "step": 98090
+    },
+    {
+      "epoch": 294.59,
+      "grad_norm": 13.00540828704834,
+      "learning_rate": 1.801801801801802e-07,
+      "loss": 0.3033,
+      "step": 98100
+    },
+    {
+      "epoch": 294.62,
+      "grad_norm": 11.652325630187988,
+      "learning_rate": 1.7917917917917919e-07,
+      "loss": 0.3295,
+      "step": 98110
+    },
+    {
+      "epoch": 294.65,
+      "grad_norm": 16.760793685913086,
+      "learning_rate": 1.7817817817817818e-07,
+      "loss": 0.3227,
+      "step": 98120
+    },
+    {
+      "epoch": 294.68,
+      "grad_norm": 15.170023918151855,
+      "learning_rate": 1.7717717717717717e-07,
+      "loss": 0.3247,
+      "step": 98130
+    },
+    {
+      "epoch": 294.71,
+      "grad_norm": 15.814078330993652,
+      "learning_rate": 1.7617617617617621e-07,
+      "loss": 0.3455,
+      "step": 98140
+    },
+    {
+      "epoch": 294.74,
+      "grad_norm": 16.483320236206055,
+      "learning_rate": 1.751751751751752e-07,
+      "loss": 0.3323,
+      "step": 98150
+    },
+    {
+      "epoch": 294.77,
+      "grad_norm": 14.266125679016113,
+      "learning_rate": 1.741741741741742e-07,
+      "loss": 0.3106,
+      "step": 98160
+    },
+    {
+      "epoch": 294.8,
+      "grad_norm": 12.863424301147461,
+      "learning_rate": 1.731731731731732e-07,
+      "loss": 0.3186,
+      "step": 98170
+    },
+    {
+      "epoch": 294.83,
+      "grad_norm": 12.47504711151123,
+      "learning_rate": 1.7217217217217218e-07,
+      "loss": 0.3453,
+      "step": 98180
+    },
+    {
+      "epoch": 294.86,
+      "grad_norm": 11.777057647705078,
+      "learning_rate": 1.7117117117117117e-07,
+      "loss": 0.2996,
+      "step": 98190
+    },
+    {
+      "epoch": 294.89,
+      "grad_norm": 15.461119651794434,
+      "learning_rate": 1.701701701701702e-07,
+      "loss": 0.3763,
+      "step": 98200
+    },
+    {
+      "epoch": 294.92,
+      "grad_norm": 25.174449920654297,
+      "learning_rate": 1.6916916916916918e-07,
+      "loss": 0.3278,
+      "step": 98210
+    },
+    {
+      "epoch": 294.95,
+      "grad_norm": 20.77471160888672,
+      "learning_rate": 1.6816816816816818e-07,
+      "loss": 0.3356,
+      "step": 98220
+    },
+    {
+      "epoch": 294.98,
+      "grad_norm": 16.06532096862793,
+      "learning_rate": 1.671671671671672e-07,
+      "loss": 0.3277,
+      "step": 98230
+    },
+    {
+      "epoch": 295.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.549491286277771,
+      "eval_runtime": 12.588,
+      "eval_samples_per_second": 794.41,
+      "eval_steps_per_second": 3.178,
+      "step": 98235
+    },
+    {
+      "epoch": 295.02,
+      "grad_norm": 12.393390655517578,
+      "learning_rate": 1.6616616616616619e-07,
+      "loss": 0.3121,
+      "step": 98240
+    },
+    {
+      "epoch": 295.05,
+      "grad_norm": 22.760889053344727,
+      "learning_rate": 1.6516516516516518e-07,
+      "loss": 0.3075,
+      "step": 98250
+    },
+    {
+      "epoch": 295.08,
+      "grad_norm": 24.913705825805664,
+      "learning_rate": 1.6416416416416417e-07,
+      "loss": 0.3007,
+      "step": 98260
+    },
+    {
+      "epoch": 295.11,
+      "grad_norm": 11.423727989196777,
+      "learning_rate": 1.631631631631632e-07,
+      "loss": 0.4049,
+      "step": 98270
+    },
+    {
+      "epoch": 295.14,
+      "grad_norm": 14.31965160369873,
+      "learning_rate": 1.6216216216216218e-07,
+      "loss": 0.3075,
+      "step": 98280
+    },
+    {
+      "epoch": 295.17,
+      "grad_norm": 10.783029556274414,
+      "learning_rate": 1.6116116116116117e-07,
+      "loss": 0.3373,
+      "step": 98290
+    },
+    {
+      "epoch": 295.2,
+      "grad_norm": 14.645768165588379,
+      "learning_rate": 1.6016016016016016e-07,
+      "loss": 0.323,
+      "step": 98300
+    },
+    {
+      "epoch": 295.23,
+      "grad_norm": 20.62546157836914,
+      "learning_rate": 1.5915915915915916e-07,
+      "loss": 0.3345,
+      "step": 98310
+    },
+    {
+      "epoch": 295.26,
+      "grad_norm": 12.165386199951172,
+      "learning_rate": 1.5815815815815815e-07,
+      "loss": 0.3196,
+      "step": 98320
+    },
+    {
+      "epoch": 295.29,
+      "grad_norm": 21.39295768737793,
+      "learning_rate": 1.5715715715715717e-07,
+      "loss": 0.3485,
+      "step": 98330
+    },
+    {
+      "epoch": 295.32,
+      "grad_norm": 20.392257690429688,
+      "learning_rate": 1.5615615615615616e-07,
+      "loss": 0.323,
+      "step": 98340
+    },
+    {
+      "epoch": 295.35,
+      "grad_norm": 13.17255973815918,
+      "learning_rate": 1.5515515515515518e-07,
+      "loss": 0.3309,
+      "step": 98350
+    },
+    {
+      "epoch": 295.38,
+      "grad_norm": 19.016233444213867,
+      "learning_rate": 1.5415415415415417e-07,
+      "loss": 0.3168,
+      "step": 98360
+    },
+    {
+      "epoch": 295.41,
+      "grad_norm": 15.294814109802246,
+      "learning_rate": 1.5315315315315316e-07,
+      "loss": 0.3398,
+      "step": 98370
+    },
+    {
+      "epoch": 295.44,
+      "grad_norm": 9.539815902709961,
+      "learning_rate": 1.5215215215215215e-07,
+      "loss": 0.2728,
+      "step": 98380
+    },
+    {
+      "epoch": 295.47,
+      "grad_norm": 19.873090744018555,
+      "learning_rate": 1.5115115115115117e-07,
+      "loss": 0.3024,
+      "step": 98390
+    },
+    {
+      "epoch": 295.5,
+      "grad_norm": 24.221525192260742,
+      "learning_rate": 1.5015015015015016e-07,
+      "loss": 0.3377,
+      "step": 98400
+    },
+    {
+      "epoch": 295.53,
+      "grad_norm": 18.842575073242188,
+      "learning_rate": 1.4914914914914915e-07,
+      "loss": 0.3313,
+      "step": 98410
+    },
+    {
+      "epoch": 295.56,
+      "grad_norm": 18.219629287719727,
+      "learning_rate": 1.4814814814814817e-07,
+      "loss": 0.3278,
+      "step": 98420
+    },
+    {
+      "epoch": 295.59,
+      "grad_norm": 25.514957427978516,
+      "learning_rate": 1.4714714714714716e-07,
+      "loss": 0.3651,
+      "step": 98430
+    },
+    {
+      "epoch": 295.62,
+      "grad_norm": 34.533790588378906,
+      "learning_rate": 1.4614614614614616e-07,
+      "loss": 0.2731,
+      "step": 98440
+    },
+    {
+      "epoch": 295.65,
+      "grad_norm": 14.030984878540039,
+      "learning_rate": 1.4514514514514517e-07,
+      "loss": 0.302,
+      "step": 98450
+    },
+    {
+      "epoch": 295.68,
+      "grad_norm": 17.93936538696289,
+      "learning_rate": 1.4414414414414417e-07,
+      "loss": 0.2964,
+      "step": 98460
+    },
+    {
+      "epoch": 295.71,
+      "grad_norm": 19.991405487060547,
+      "learning_rate": 1.4314314314314316e-07,
+      "loss": 0.3397,
+      "step": 98470
+    },
+    {
+      "epoch": 295.74,
+      "grad_norm": 13.768113136291504,
+      "learning_rate": 1.4214214214214215e-07,
+      "loss": 0.3509,
+      "step": 98480
+    },
+    {
+      "epoch": 295.77,
+      "grad_norm": 14.257954597473145,
+      "learning_rate": 1.4114114114114117e-07,
+      "loss": 0.3422,
+      "step": 98490
+    },
+    {
+      "epoch": 295.8,
+      "grad_norm": 12.76844596862793,
+      "learning_rate": 1.4014014014014016e-07,
+      "loss": 0.3072,
+      "step": 98500
+    },
+    {
+      "epoch": 295.83,
+      "grad_norm": 23.93326759338379,
+      "learning_rate": 1.3913913913913915e-07,
+      "loss": 0.3594,
+      "step": 98510
+    },
+    {
+      "epoch": 295.86,
+      "grad_norm": 19.45301628112793,
+      "learning_rate": 1.3813813813813814e-07,
+      "loss": 0.3398,
+      "step": 98520
+    },
+    {
+      "epoch": 295.89,
+      "grad_norm": 13.732878684997559,
+      "learning_rate": 1.3713713713713716e-07,
+      "loss": 0.3515,
+      "step": 98530
+    },
+    {
+      "epoch": 295.92,
+      "grad_norm": 19.06048583984375,
+      "learning_rate": 1.3613613613613615e-07,
+      "loss": 0.296,
+      "step": 98540
+    },
+    {
+      "epoch": 295.95,
+      "grad_norm": 18.801651000976562,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 0.2878,
+      "step": 98550
+    },
+    {
+      "epoch": 295.98,
+      "grad_norm": 11.700456619262695,
+      "learning_rate": 1.3413413413413414e-07,
+      "loss": 0.3115,
+      "step": 98560
+    },
+    {
+      "epoch": 296.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.550277829170227,
+      "eval_runtime": 12.7135,
+      "eval_samples_per_second": 786.568,
+      "eval_steps_per_second": 3.146,
+      "step": 98568
+    },
+    {
+      "epoch": 296.01,
+      "grad_norm": 12.897662162780762,
+      "learning_rate": 1.3313313313313313e-07,
+      "loss": 0.2411,
+      "step": 98570
+    },
+    {
+      "epoch": 296.04,
+      "grad_norm": 14.836771011352539,
+      "learning_rate": 1.3213213213213215e-07,
+      "loss": 0.2761,
+      "step": 98580
+    },
+    {
+      "epoch": 296.07,
+      "grad_norm": 15.093950271606445,
+      "learning_rate": 1.3113113113113114e-07,
+      "loss": 0.3527,
+      "step": 98590
+    },
+    {
+      "epoch": 296.1,
+      "grad_norm": 15.355332374572754,
+      "learning_rate": 1.3013013013013013e-07,
+      "loss": 0.3512,
+      "step": 98600
+    },
+    {
+      "epoch": 296.13,
+      "grad_norm": 8.834267616271973,
+      "learning_rate": 1.2912912912912912e-07,
+      "loss": 0.3216,
+      "step": 98610
+    },
+    {
+      "epoch": 296.16,
+      "grad_norm": 17.48318862915039,
+      "learning_rate": 1.2812812812812814e-07,
+      "loss": 0.3062,
+      "step": 98620
+    },
+    {
+      "epoch": 296.19,
+      "grad_norm": 21.506179809570312,
+      "learning_rate": 1.2712712712712713e-07,
+      "loss": 0.3579,
+      "step": 98630
+    },
+    {
+      "epoch": 296.22,
+      "grad_norm": 9.43555736541748,
+      "learning_rate": 1.2612612612612613e-07,
+      "loss": 0.2765,
+      "step": 98640
+    },
+    {
+      "epoch": 296.25,
+      "grad_norm": 23.990652084350586,
+      "learning_rate": 1.2512512512512514e-07,
+      "loss": 0.3475,
+      "step": 98650
+    },
+    {
+      "epoch": 296.28,
+      "grad_norm": 12.934406280517578,
+      "learning_rate": 1.2412412412412414e-07,
+      "loss": 0.3156,
+      "step": 98660
+    },
+    {
+      "epoch": 296.31,
+      "grad_norm": 18.725576400756836,
+      "learning_rate": 1.2312312312312313e-07,
+      "loss": 0.2834,
+      "step": 98670
+    },
+    {
+      "epoch": 296.34,
+      "grad_norm": 13.866150856018066,
+      "learning_rate": 1.2212212212212212e-07,
+      "loss": 0.3561,
+      "step": 98680
+    },
+    {
+      "epoch": 296.37,
+      "grad_norm": 17.21200942993164,
+      "learning_rate": 1.2112112112112114e-07,
+      "loss": 0.3094,
+      "step": 98690
+    },
+    {
+      "epoch": 296.4,
+      "grad_norm": 19.460330963134766,
+      "learning_rate": 1.2012012012012013e-07,
+      "loss": 0.3462,
+      "step": 98700
+    },
+    {
+      "epoch": 296.43,
+      "grad_norm": 19.61447525024414,
+      "learning_rate": 1.1911911911911912e-07,
+      "loss": 0.3103,
+      "step": 98710
+    },
+    {
+      "epoch": 296.46,
+      "grad_norm": 14.759013175964355,
+      "learning_rate": 1.1811811811811814e-07,
+      "loss": 0.342,
+      "step": 98720
+    },
+    {
+      "epoch": 296.49,
+      "grad_norm": 14.679244041442871,
+      "learning_rate": 1.1711711711711713e-07,
+      "loss": 0.2911,
+      "step": 98730
+    },
+    {
+      "epoch": 296.52,
+      "grad_norm": 13.90548324584961,
+      "learning_rate": 1.1611611611611612e-07,
+      "loss": 0.3552,
+      "step": 98740
+    },
+    {
+      "epoch": 296.55,
+      "grad_norm": 10.689663887023926,
+      "learning_rate": 1.1511511511511513e-07,
+      "loss": 0.2929,
+      "step": 98750
+    },
+    {
+      "epoch": 296.58,
+      "grad_norm": 15.175422668457031,
+      "learning_rate": 1.1411411411411412e-07,
+      "loss": 0.333,
+      "step": 98760
+    },
+    {
+      "epoch": 296.61,
+      "grad_norm": 16.35551643371582,
+      "learning_rate": 1.1311311311311313e-07,
+      "loss": 0.3389,
+      "step": 98770
+    },
+    {
+      "epoch": 296.64,
+      "grad_norm": 17.737401962280273,
+      "learning_rate": 1.1211211211211212e-07,
+      "loss": 0.3139,
+      "step": 98780
+    },
+    {
+      "epoch": 296.67,
+      "grad_norm": 18.34389305114746,
+      "learning_rate": 1.1111111111111112e-07,
+      "loss": 0.3433,
+      "step": 98790
+    },
+    {
+      "epoch": 296.7,
+      "grad_norm": 13.155138969421387,
+      "learning_rate": 1.1011011011011011e-07,
+      "loss": 0.3208,
+      "step": 98800
+    },
+    {
+      "epoch": 296.73,
+      "grad_norm": 9.929429054260254,
+      "learning_rate": 1.091091091091091e-07,
+      "loss": 0.3566,
+      "step": 98810
+    },
+    {
+      "epoch": 296.76,
+      "grad_norm": 25.24324607849121,
+      "learning_rate": 1.0810810810810812e-07,
+      "loss": 0.3202,
+      "step": 98820
+    },
+    {
+      "epoch": 296.79,
+      "grad_norm": 16.974205017089844,
+      "learning_rate": 1.0710710710710712e-07,
+      "loss": 0.3513,
+      "step": 98830
+    },
+    {
+      "epoch": 296.82,
+      "grad_norm": 12.666238784790039,
+      "learning_rate": 1.0610610610610611e-07,
+      "loss": 0.329,
+      "step": 98840
+    },
+    {
+      "epoch": 296.85,
+      "grad_norm": 19.592561721801758,
+      "learning_rate": 1.0510510510510513e-07,
+      "loss": 0.3084,
+      "step": 98850
+    },
+    {
+      "epoch": 296.88,
+      "grad_norm": 13.523087501525879,
+      "learning_rate": 1.0410410410410412e-07,
+      "loss": 0.3309,
+      "step": 98860
+    },
+    {
+      "epoch": 296.91,
+      "grad_norm": 14.393192291259766,
+      "learning_rate": 1.0310310310310311e-07,
+      "loss": 0.3178,
+      "step": 98870
+    },
+    {
+      "epoch": 296.94,
+      "grad_norm": 13.908355712890625,
+      "learning_rate": 1.021021021021021e-07,
+      "loss": 0.3073,
+      "step": 98880
+    },
+    {
+      "epoch": 296.97,
+      "grad_norm": 13.29915714263916,
+      "learning_rate": 1.0110110110110112e-07,
+      "loss": 0.3239,
+      "step": 98890
+    },
+    {
+      "epoch": 297.0,
+      "grad_norm": 14.719878196716309,
+      "learning_rate": 1.0010010010010011e-07,
+      "loss": 0.3103,
+      "step": 98900
+    },
+    {
+      "epoch": 297.0,
+      "eval_accuracy": 0.8643,
+      "eval_loss": 0.5503737926483154,
+      "eval_runtime": 12.642,
+      "eval_samples_per_second": 791.015,
+      "eval_steps_per_second": 3.164,
+      "step": 98901
+    },
+    {
+      "epoch": 297.03,
+      "grad_norm": 17.77787208557129,
+      "learning_rate": 9.90990990990991e-08,
+      "loss": 0.3685,
+      "step": 98910
+    },
+    {
+      "epoch": 297.06,
+      "grad_norm": 17.867164611816406,
+      "learning_rate": 9.809809809809811e-08,
+      "loss": 0.3,
+      "step": 98920
+    },
+    {
+      "epoch": 297.09,
+      "grad_norm": 21.947612762451172,
+      "learning_rate": 9.70970970970971e-08,
+      "loss": 0.3266,
+      "step": 98930
+    },
+    {
+      "epoch": 297.12,
+      "grad_norm": 12.852577209472656,
+      "learning_rate": 9.60960960960961e-08,
+      "loss": 0.3499,
+      "step": 98940
+    },
+    {
+      "epoch": 297.15,
+      "grad_norm": 12.509542465209961,
+      "learning_rate": 9.509509509509511e-08,
+      "loss": 0.3566,
+      "step": 98950
+    },
+    {
+      "epoch": 297.18,
+      "grad_norm": 13.387910842895508,
+      "learning_rate": 9.40940940940941e-08,
+      "loss": 0.2535,
+      "step": 98960
+    },
+    {
+      "epoch": 297.21,
+      "grad_norm": 15.900568008422852,
+      "learning_rate": 9.30930930930931e-08,
+      "loss": 0.3819,
+      "step": 98970
+    },
+    {
+      "epoch": 297.24,
+      "grad_norm": 17.37117576599121,
+      "learning_rate": 9.209209209209209e-08,
+      "loss": 0.3484,
+      "step": 98980
+    },
+    {
+      "epoch": 297.27,
+      "grad_norm": 17.52071189880371,
+      "learning_rate": 9.10910910910911e-08,
+      "loss": 0.3347,
+      "step": 98990
+    },
+    {
+      "epoch": 297.3,
+      "grad_norm": 25.807437896728516,
+      "learning_rate": 9.00900900900901e-08,
+      "loss": 0.3566,
+      "step": 99000
+    },
+    {
+      "epoch": 297.33,
+      "grad_norm": 19.961111068725586,
+      "learning_rate": 8.908908908908909e-08,
+      "loss": 0.3797,
+      "step": 99010
+    },
+    {
+      "epoch": 297.36,
+      "grad_norm": 10.74624252319336,
+      "learning_rate": 8.808808808808811e-08,
+      "loss": 0.3149,
+      "step": 99020
+    },
+    {
+      "epoch": 297.39,
+      "grad_norm": 16.93206024169922,
+      "learning_rate": 8.70870870870871e-08,
+      "loss": 0.3322,
+      "step": 99030
+    },
+    {
+      "epoch": 297.42,
+      "grad_norm": 17.694971084594727,
+      "learning_rate": 8.608608608608609e-08,
+      "loss": 0.2929,
+      "step": 99040
+    },
+    {
+      "epoch": 297.45,
+      "grad_norm": 10.793231964111328,
+      "learning_rate": 8.50850850850851e-08,
+      "loss": 0.3316,
+      "step": 99050
+    },
+    {
+      "epoch": 297.48,
+      "grad_norm": 14.06987476348877,
+      "learning_rate": 8.408408408408409e-08,
+      "loss": 0.387,
+      "step": 99060
+    },
+    {
+      "epoch": 297.51,
+      "grad_norm": 13.655963897705078,
+      "learning_rate": 8.308308308308309e-08,
+      "loss": 0.3422,
+      "step": 99070
+    },
+    {
+      "epoch": 297.54,
+      "grad_norm": 15.021867752075195,
+      "learning_rate": 8.208208208208208e-08,
+      "loss": 0.3402,
+      "step": 99080
+    },
+    {
+      "epoch": 297.57,
+      "grad_norm": 21.111528396606445,
+      "learning_rate": 8.108108108108109e-08,
+      "loss": 0.4457,
+      "step": 99090
+    },
+    {
+      "epoch": 297.6,
+      "grad_norm": 14.999329566955566,
+      "learning_rate": 8.008008008008008e-08,
+      "loss": 0.3179,
+      "step": 99100
+    },
+    {
+      "epoch": 297.63,
+      "grad_norm": 12.696755409240723,
+      "learning_rate": 7.907907907907907e-08,
+      "loss": 0.3416,
+      "step": 99110
+    },
+    {
+      "epoch": 297.66,
+      "grad_norm": 24.10008430480957,
+      "learning_rate": 7.807807807807808e-08,
+      "loss": 0.3175,
+      "step": 99120
+    },
+    {
+      "epoch": 297.69,
+      "grad_norm": 9.3154935836792,
+      "learning_rate": 7.707707707707708e-08,
+      "loss": 0.3013,
+      "step": 99130
+    },
+    {
+      "epoch": 297.72,
+      "grad_norm": 14.79358959197998,
+      "learning_rate": 7.607607607607608e-08,
+      "loss": 0.3281,
+      "step": 99140
+    },
+    {
+      "epoch": 297.75,
+      "grad_norm": 14.416643142700195,
+      "learning_rate": 7.507507507507508e-08,
+      "loss": 0.3393,
+      "step": 99150
+    },
+    {
+      "epoch": 297.78,
+      "grad_norm": 14.998799324035645,
+      "learning_rate": 7.407407407407409e-08,
+      "loss": 0.2582,
+      "step": 99160
+    },
+    {
+      "epoch": 297.81,
+      "grad_norm": 19.169893264770508,
+      "learning_rate": 7.307307307307308e-08,
+      "loss": 0.2765,
+      "step": 99170
+    },
+    {
+      "epoch": 297.84,
+      "grad_norm": 14.578001022338867,
+      "learning_rate": 7.207207207207208e-08,
+      "loss": 0.3094,
+      "step": 99180
+    },
+    {
+      "epoch": 297.87,
+      "grad_norm": 16.767444610595703,
+      "learning_rate": 7.107107107107107e-08,
+      "loss": 0.3379,
+      "step": 99190
+    },
+    {
+      "epoch": 297.9,
+      "grad_norm": 16.213525772094727,
+      "learning_rate": 7.007007007007008e-08,
+      "loss": 0.3817,
+      "step": 99200
+    },
+    {
+      "epoch": 297.93,
+      "grad_norm": 14.708305358886719,
+      "learning_rate": 6.906906906906907e-08,
+      "loss": 0.4031,
+      "step": 99210
+    },
+    {
+      "epoch": 297.96,
+      "grad_norm": 11.830233573913574,
+      "learning_rate": 6.806806806806808e-08,
+      "loss": 0.3318,
+      "step": 99220
+    },
+    {
+      "epoch": 297.99,
+      "grad_norm": 16.058086395263672,
+      "learning_rate": 6.706706706706707e-08,
+      "loss": 0.2616,
+      "step": 99230
+    },
+    {
+      "epoch": 298.0,
+      "eval_accuracy": 0.8648,
+      "eval_loss": 0.5507569313049316,
+      "eval_runtime": 12.9128,
+      "eval_samples_per_second": 774.423,
+      "eval_steps_per_second": 3.098,
+      "step": 99234
+    },
+    {
+      "epoch": 298.02,
+      "grad_norm": 15.163691520690918,
+      "learning_rate": 6.606606606606607e-08,
+      "loss": 0.3171,
+      "step": 99240
+    },
+    {
+      "epoch": 298.05,
+      "grad_norm": 13.903767585754395,
+      "learning_rate": 6.506506506506507e-08,
+      "loss": 0.3293,
+      "step": 99250
+    },
+    {
+      "epoch": 298.08,
+      "grad_norm": 16.759506225585938,
+      "learning_rate": 6.406406406406407e-08,
+      "loss": 0.2896,
+      "step": 99260
+    },
+    {
+      "epoch": 298.11,
+      "grad_norm": 15.209610939025879,
+      "learning_rate": 6.306306306306306e-08,
+      "loss": 0.3143,
+      "step": 99270
+    },
+    {
+      "epoch": 298.14,
+      "grad_norm": 12.948027610778809,
+      "learning_rate": 6.206206206206207e-08,
+      "loss": 0.3604,
+      "step": 99280
+    },
+    {
+      "epoch": 298.17,
+      "grad_norm": 19.160259246826172,
+      "learning_rate": 6.106106106106106e-08,
+      "loss": 0.3116,
+      "step": 99290
+    },
+    {
+      "epoch": 298.2,
+      "grad_norm": 13.701067924499512,
+      "learning_rate": 6.006006006006006e-08,
+      "loss": 0.3524,
+      "step": 99300
+    },
+    {
+      "epoch": 298.23,
+      "grad_norm": 9.308465003967285,
+      "learning_rate": 5.905905905905907e-08,
+      "loss": 0.3058,
+      "step": 99310
+    },
+    {
+      "epoch": 298.26,
+      "grad_norm": 16.943267822265625,
+      "learning_rate": 5.805805805805806e-08,
+      "loss": 0.3111,
+      "step": 99320
+    },
+    {
+      "epoch": 298.29,
+      "grad_norm": 17.831790924072266,
+      "learning_rate": 5.705705705705706e-08,
+      "loss": 0.29,
+      "step": 99330
+    },
+    {
+      "epoch": 298.32,
+      "grad_norm": 25.99674415588379,
+      "learning_rate": 5.605605605605606e-08,
+      "loss": 0.2627,
+      "step": 99340
+    },
+    {
+      "epoch": 298.35,
+      "grad_norm": 36.646724700927734,
+      "learning_rate": 5.505505505505506e-08,
+      "loss": 0.3249,
+      "step": 99350
+    },
+    {
+      "epoch": 298.38,
+      "grad_norm": 14.348864555358887,
+      "learning_rate": 5.405405405405406e-08,
+      "loss": 0.2596,
+      "step": 99360
+    },
+    {
+      "epoch": 298.41,
+      "grad_norm": 13.356197357177734,
+      "learning_rate": 5.3053053053053054e-08,
+      "loss": 0.2734,
+      "step": 99370
+    },
+    {
+      "epoch": 298.44,
+      "grad_norm": 12.323141098022461,
+      "learning_rate": 5.205205205205206e-08,
+      "loss": 0.2909,
+      "step": 99380
+    },
+    {
+      "epoch": 298.47,
+      "grad_norm": 17.2496337890625,
+      "learning_rate": 5.105105105105105e-08,
+      "loss": 0.2895,
+      "step": 99390
+    },
+    {
+      "epoch": 298.5,
+      "grad_norm": 17.626937866210938,
+      "learning_rate": 5.0050050050050056e-08,
+      "loss": 0.2929,
+      "step": 99400
+    },
+    {
+      "epoch": 298.53,
+      "grad_norm": 19.492650985717773,
+      "learning_rate": 4.9049049049049055e-08,
+      "loss": 0.3414,
+      "step": 99410
+    },
+    {
+      "epoch": 298.56,
+      "grad_norm": 20.08698081970215,
+      "learning_rate": 4.804804804804805e-08,
+      "loss": 0.3048,
+      "step": 99420
+    },
+    {
+      "epoch": 298.59,
+      "grad_norm": 14.274243354797363,
+      "learning_rate": 4.704704704704705e-08,
+      "loss": 0.2728,
+      "step": 99430
+    },
+    {
+      "epoch": 298.62,
+      "grad_norm": 13.778084754943848,
+      "learning_rate": 4.6046046046046043e-08,
+      "loss": 0.2823,
+      "step": 99440
+    },
+    {
+      "epoch": 298.65,
+      "grad_norm": 15.873741149902344,
+      "learning_rate": 4.504504504504505e-08,
+      "loss": 0.3476,
+      "step": 99450
+    },
+    {
+      "epoch": 298.68,
+      "grad_norm": 16.204940795898438,
+      "learning_rate": 4.4044044044044054e-08,
+      "loss": 0.2704,
+      "step": 99460
+    },
+    {
+      "epoch": 298.71,
+      "grad_norm": 18.95944595336914,
+      "learning_rate": 4.3043043043043045e-08,
+      "loss": 0.3714,
+      "step": 99470
+    },
+    {
+      "epoch": 298.74,
+      "grad_norm": 11.249991416931152,
+      "learning_rate": 4.2042042042042044e-08,
+      "loss": 0.3214,
+      "step": 99480
+    },
+    {
+      "epoch": 298.77,
+      "grad_norm": 20.00392723083496,
+      "learning_rate": 4.104104104104104e-08,
+      "loss": 0.2661,
+      "step": 99490
+    },
+    {
+      "epoch": 298.8,
+      "grad_norm": 22.023340225219727,
+      "learning_rate": 4.004004004004004e-08,
+      "loss": 0.38,
+      "step": 99500
+    },
+    {
+      "epoch": 298.83,
+      "grad_norm": 20.28693199157715,
+      "learning_rate": 3.903903903903904e-08,
+      "loss": 0.2986,
+      "step": 99510
+    },
+    {
+      "epoch": 298.86,
+      "grad_norm": 17.731781005859375,
+      "learning_rate": 3.803803803803804e-08,
+      "loss": 0.333,
+      "step": 99520
+    },
+    {
+      "epoch": 298.89,
+      "grad_norm": 17.1861629486084,
+      "learning_rate": 3.703703703703704e-08,
+      "loss": 0.2861,
+      "step": 99530
+    },
+    {
+      "epoch": 298.92,
+      "grad_norm": 12.45194149017334,
+      "learning_rate": 3.603603603603604e-08,
+      "loss": 0.2993,
+      "step": 99540
+    },
+    {
+      "epoch": 298.95,
+      "grad_norm": 13.979104995727539,
+      "learning_rate": 3.503503503503504e-08,
+      "loss": 0.3108,
+      "step": 99550
+    },
+    {
+      "epoch": 298.98,
+      "grad_norm": 12.693013191223145,
+      "learning_rate": 3.403403403403404e-08,
+      "loss": 0.3072,
+      "step": 99560
+    },
+    {
+      "epoch": 299.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.550467848777771,
+      "eval_runtime": 12.6448,
+      "eval_samples_per_second": 790.842,
+      "eval_steps_per_second": 3.163,
+      "step": 99567
+    },
+    {
+      "epoch": 299.01,
+      "grad_norm": 15.526609420776367,
+      "learning_rate": 3.303303303303304e-08,
+      "loss": 0.2893,
+      "step": 99570
+    },
+    {
+      "epoch": 299.04,
+      "grad_norm": 22.6369571685791,
+      "learning_rate": 3.2032032032032035e-08,
+      "loss": 0.3328,
+      "step": 99580
+    },
+    {
+      "epoch": 299.07,
+      "grad_norm": 13.886428833007812,
+      "learning_rate": 3.1031031031031034e-08,
+      "loss": 0.283,
+      "step": 99590
+    },
+    {
+      "epoch": 299.1,
+      "grad_norm": 15.69870662689209,
+      "learning_rate": 3.003003003003003e-08,
+      "loss": 0.2918,
+      "step": 99600
+    },
+    {
+      "epoch": 299.13,
+      "grad_norm": 16.28816795349121,
+      "learning_rate": 2.902902902902903e-08,
+      "loss": 0.3095,
+      "step": 99610
+    },
+    {
+      "epoch": 299.16,
+      "grad_norm": 18.956165313720703,
+      "learning_rate": 2.802802802802803e-08,
+      "loss": 0.3042,
+      "step": 99620
+    },
+    {
+      "epoch": 299.19,
+      "grad_norm": 17.06728744506836,
+      "learning_rate": 2.702702702702703e-08,
+      "loss": 0.2904,
+      "step": 99630
+    },
+    {
+      "epoch": 299.22,
+      "grad_norm": 12.715773582458496,
+      "learning_rate": 2.602602602602603e-08,
+      "loss": 0.34,
+      "step": 99640
+    },
+    {
+      "epoch": 299.25,
+      "grad_norm": 13.41841983795166,
+      "learning_rate": 2.5025025025025028e-08,
+      "loss": 0.305,
+      "step": 99650
+    },
+    {
+      "epoch": 299.28,
+      "grad_norm": 22.93665885925293,
+      "learning_rate": 2.4024024024024027e-08,
+      "loss": 0.3829,
+      "step": 99660
+    },
+    {
+      "epoch": 299.31,
+      "grad_norm": 14.447463035583496,
+      "learning_rate": 2.3023023023023022e-08,
+      "loss": 0.3015,
+      "step": 99670
+    },
+    {
+      "epoch": 299.34,
+      "grad_norm": 10.823923110961914,
+      "learning_rate": 2.2022022022022027e-08,
+      "loss": 0.3355,
+      "step": 99680
+    },
+    {
+      "epoch": 299.37,
+      "grad_norm": 18.890472412109375,
+      "learning_rate": 2.1021021021021022e-08,
+      "loss": 0.323,
+      "step": 99690
+    },
+    {
+      "epoch": 299.4,
+      "grad_norm": 16.06194305419922,
+      "learning_rate": 2.002002002002002e-08,
+      "loss": 0.2684,
+      "step": 99700
+    },
+    {
+      "epoch": 299.43,
+      "grad_norm": 14.057879447937012,
+      "learning_rate": 1.901901901901902e-08,
+      "loss": 0.3368,
+      "step": 99710
+    },
+    {
+      "epoch": 299.46,
+      "grad_norm": 14.292717933654785,
+      "learning_rate": 1.801801801801802e-08,
+      "loss": 0.3006,
+      "step": 99720
+    },
+    {
+      "epoch": 299.49,
+      "grad_norm": 15.200785636901855,
+      "learning_rate": 1.701701701701702e-08,
+      "loss": 0.3173,
+      "step": 99730
+    },
+    {
+      "epoch": 299.52,
+      "grad_norm": 12.914721488952637,
+      "learning_rate": 1.6016016016016018e-08,
+      "loss": 0.3361,
+      "step": 99740
+    },
+    {
+      "epoch": 299.55,
+      "grad_norm": 12.097268104553223,
+      "learning_rate": 1.5015015015015016e-08,
+      "loss": 0.3164,
+      "step": 99750
+    },
+    {
+      "epoch": 299.58,
+      "grad_norm": 12.614750862121582,
+      "learning_rate": 1.4014014014014015e-08,
+      "loss": 0.2734,
+      "step": 99760
+    },
+    {
+      "epoch": 299.61,
+      "grad_norm": 20.287817001342773,
+      "learning_rate": 1.3013013013013015e-08,
+      "loss": 0.3131,
+      "step": 99770
+    },
+    {
+      "epoch": 299.64,
+      "grad_norm": 17.952526092529297,
+      "learning_rate": 1.2012012012012013e-08,
+      "loss": 0.3025,
+      "step": 99780
+    },
+    {
+      "epoch": 299.67,
+      "grad_norm": 11.732912063598633,
+      "learning_rate": 1.1011011011011013e-08,
+      "loss": 0.3261,
+      "step": 99790
+    },
+    {
+      "epoch": 299.7,
+      "grad_norm": 35.14999771118164,
+      "learning_rate": 1.001001001001001e-08,
+      "loss": 0.3162,
+      "step": 99800
+    },
+    {
+      "epoch": 299.73,
+      "grad_norm": 20.89388084411621,
+      "learning_rate": 9.00900900900901e-09,
+      "loss": 0.3347,
+      "step": 99810
+    },
+    {
+      "epoch": 299.76,
+      "grad_norm": 12.35584831237793,
+      "learning_rate": 8.008008008008009e-09,
+      "loss": 0.2852,
+      "step": 99820
+    },
+    {
+      "epoch": 299.79,
+      "grad_norm": 11.762072563171387,
+      "learning_rate": 7.007007007007007e-09,
+      "loss": 0.3537,
+      "step": 99830
+    },
+    {
+      "epoch": 299.82,
+      "grad_norm": 11.840642929077148,
+      "learning_rate": 6.006006006006007e-09,
+      "loss": 0.2847,
+      "step": 99840
+    },
+    {
+      "epoch": 299.85,
+      "grad_norm": 15.550333976745605,
+      "learning_rate": 5.005005005005005e-09,
+      "loss": 0.3409,
+      "step": 99850
+    },
+    {
+      "epoch": 299.88,
+      "grad_norm": 19.057191848754883,
+      "learning_rate": 4.0040040040040044e-09,
+      "loss": 0.3416,
+      "step": 99860
+    },
+    {
+      "epoch": 299.91,
+      "grad_norm": 17.614805221557617,
+      "learning_rate": 3.0030030030030033e-09,
+      "loss": 0.3971,
+      "step": 99870
+    },
+    {
+      "epoch": 299.94,
+      "grad_norm": 16.798704147338867,
+      "learning_rate": 2.0020020020020022e-09,
+      "loss": 0.3379,
+      "step": 99880
+    },
+    {
+      "epoch": 299.97,
+      "grad_norm": 14.02078628540039,
+      "learning_rate": 1.0010010010010011e-09,
+      "loss": 0.3117,
+      "step": 99890
+    },
+    {
+      "epoch": 300.0,
+      "grad_norm": 130.68911743164062,
+      "learning_rate": 0.0,
+      "loss": 0.3333,
+      "step": 99900
+    },
+    {
+      "epoch": 300.0,
+      "eval_accuracy": 0.8646,
+      "eval_loss": 0.5505151152610779,
+      "eval_runtime": 13.0,
+      "eval_samples_per_second": 769.229,
+      "eval_steps_per_second": 3.077,
+      "step": 99900
+    },
+    {
+      "epoch": 300.0,
+      "step": 99900,
+      "total_flos": 3.177816825415706e+20,
+      "train_loss": 0.4971393480458417,
+      "train_runtime": 46729.5214,
+      "train_samples_per_second": 272.847,
+      "train_steps_per_second": 2.138
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 99900,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 300,
+  "save_steps": 500,
+  "total_flos": 3.177816825415706e+20,
+  "train_batch_size": 128,
+  "trial_name": null,
+  "trial_params": null
+}