diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.5,
+  "eval_steps": 500,
+  "global_step": 250000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005,
+      "grad_norm": 55.07521057128906,
+      "learning_rate": 9.99816e-07,
+      "loss": 1.4625,
+      "step": 50
+    },
+    {
+      "epoch": 0.001,
+      "grad_norm": 96.95206451416016,
+      "learning_rate": 9.99616e-07,
+      "loss": 1.4538,
+      "step": 100
+    },
+    {
+      "epoch": 0.0015,
+      "grad_norm": 49.72062683105469,
+      "learning_rate": 9.99416e-07,
+      "loss": 1.3503,
+      "step": 150
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 62.22916793823242,
+      "learning_rate": 9.9922e-07,
+      "loss": 1.3127,
+      "step": 200
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 38.854434967041016,
+      "learning_rate": 9.99024e-07,
+      "loss": 1.261,
+      "step": 250
+    },
+    {
+      "epoch": 0.003,
+      "grad_norm": 49.72938919067383,
+      "learning_rate": 9.98824e-07,
+      "loss": 1.1819,
+      "step": 300
+    },
+    {
+      "epoch": 0.0035,
+      "grad_norm": 68.74335479736328,
+      "learning_rate": 9.986239999999999e-07,
+      "loss": 1.1701,
+      "step": 350
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 51.449588775634766,
+      "learning_rate": 9.984239999999999e-07,
+      "loss": 1.1705,
+      "step": 400
+    },
+    {
+      "epoch": 0.0045,
+      "grad_norm": 42.585655212402344,
+      "learning_rate": 9.98224e-07,
+      "loss": 1.0597,
+      "step": 450
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 62.51144027709961,
+      "learning_rate": 9.98024e-07,
+      "loss": 1.0494,
+      "step": 500
+    },
+    {
+      "epoch": 0.0055,
+      "grad_norm": 40.80481719970703,
+      "learning_rate": 9.97824e-07,
+      "loss": 1.027,
+      "step": 550
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 63.342613220214844,
+      "learning_rate": 9.97628e-07,
+      "loss": 1.048,
+      "step": 600
+    },
+    {
+      "epoch": 0.0065,
+      "grad_norm": 38.498329162597656,
+      "learning_rate": 9.97428e-07,
+      "loss": 0.9951,
+      "step": 650
+    },
+    {
+      "epoch": 0.007,
+      "grad_norm": 25.70156478881836,
+      "learning_rate": 9.972279999999999e-07,
+      "loss": 0.9348,
+      "step": 700
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 54.72833251953125,
+      "learning_rate": 9.97028e-07,
+      "loss": 0.9405,
+      "step": 750
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 51.39662170410156,
+      "learning_rate": 9.96828e-07,
+      "loss": 0.9214,
+      "step": 800
+    },
+    {
+      "epoch": 0.0085,
+      "grad_norm": 57.055030822753906,
+      "learning_rate": 9.966279999999998e-07,
+      "loss": 0.9247,
+      "step": 850
+    },
+    {
+      "epoch": 0.009,
+      "grad_norm": 48.74729919433594,
+      "learning_rate": 9.96428e-07,
+      "loss": 0.9254,
+      "step": 900
+    },
+    {
+      "epoch": 0.0095,
+      "grad_norm": 69.1877212524414,
+      "learning_rate": 9.96228e-07,
+      "loss": 0.7783,
+      "step": 950
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 58.19287872314453,
+      "learning_rate": 9.96028e-07,
+      "loss": 0.9523,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0105,
+      "grad_norm": 90.27144622802734,
+      "learning_rate": 9.95828e-07,
+      "loss": 0.8884,
+      "step": 1050
+    },
+    {
+      "epoch": 0.011,
+      "grad_norm": 47.847496032714844,
+      "learning_rate": 9.95628e-07,
+      "loss": 0.765,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0115,
+      "grad_norm": 60.23532485961914,
+      "learning_rate": 9.95428e-07,
+      "loss": 0.9025,
+      "step": 1150
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 65.98821258544922,
+      "learning_rate": 9.95228e-07,
+      "loss": 0.8743,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 41.296512603759766,
+      "learning_rate": 9.95028e-07,
+      "loss": 0.7958,
+      "step": 1250
+    },
+    {
+      "epoch": 0.013,
+      "grad_norm": 10.350058555603027,
+      "learning_rate": 9.94828e-07,
+      "loss": 0.7748,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0135,
+      "grad_norm": 100.7204818725586,
+      "learning_rate": 9.94628e-07,
+      "loss": 0.7265,
+      "step": 1350
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 33.72673797607422,
+      "learning_rate": 9.944279999999999e-07,
+      "loss": 0.8833,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0145,
+      "grad_norm": 51.3131217956543,
+      "learning_rate": 9.942279999999999e-07,
+      "loss": 0.6357,
+      "step": 1450
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 62.02531814575195,
+      "learning_rate": 9.94028e-07,
+      "loss": 0.8208,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0155,
+      "grad_norm": 79.17247772216797,
+      "learning_rate": 9.93828e-07,
+      "loss": 0.7981,
+      "step": 1550
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 45.16271209716797,
+      "learning_rate": 9.936279999999998e-07,
+      "loss": 0.9424,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0165,
+      "grad_norm": 25.51656723022461,
+      "learning_rate": 9.93428e-07,
+      "loss": 0.7381,
+      "step": 1650
+    },
+    {
+      "epoch": 0.017,
+      "grad_norm": 67.2197265625,
+      "learning_rate": 9.93228e-07,
+      "loss": 0.8223,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 68.64763641357422,
+      "learning_rate": 9.93028e-07,
+      "loss": 0.8728,
+      "step": 1750
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 73.73970031738281,
+      "learning_rate": 9.92828e-07,
+      "loss": 0.8765,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0185,
+      "grad_norm": 91.64149475097656,
+      "learning_rate": 9.92628e-07,
+      "loss": 0.7413,
+      "step": 1850
+    },
+    {
+      "epoch": 0.019,
+      "grad_norm": 65.28728485107422,
+      "learning_rate": 9.92428e-07,
+      "loss": 0.8043,
+      "step": 1900
+    },
+    {
+      "epoch": 0.0195,
+      "grad_norm": 67.93745422363281,
+      "learning_rate": 9.92228e-07,
+      "loss": 0.7773,
+      "step": 1950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 59.84891128540039,
+      "learning_rate": 9.92028e-07,
+      "loss": 0.6151,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0205,
+      "grad_norm": 25.66044044494629,
+      "learning_rate": 9.91828e-07,
+      "loss": 0.8568,
+      "step": 2050
+    },
+    {
+      "epoch": 0.021,
+      "grad_norm": 88.2499008178711,
+      "learning_rate": 9.916279999999999e-07,
+      "loss": 0.8048,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0215,
+      "grad_norm": 80.02753448486328,
+      "learning_rate": 9.914279999999999e-07,
+      "loss": 0.8383,
+      "step": 2150
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 36.008663177490234,
+      "learning_rate": 9.912279999999999e-07,
+      "loss": 0.7106,
+      "step": 2200
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 19.52667999267578,
+      "learning_rate": 9.91028e-07,
+      "loss": 0.5627,
+      "step": 2250
+    },
+    {
+      "epoch": 0.023,
+      "grad_norm": 44.50703048706055,
+      "learning_rate": 9.90828e-07,
+      "loss": 0.8129,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0235,
+      "grad_norm": 31.829010009765625,
+      "learning_rate": 9.906279999999998e-07,
+      "loss": 0.5897,
+      "step": 2350
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 80.26560974121094,
+      "learning_rate": 9.90428e-07,
+      "loss": 0.7319,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0245,
+      "grad_norm": 46.38948440551758,
+      "learning_rate": 9.90228e-07,
+      "loss": 0.8328,
+      "step": 2450
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 36.038612365722656,
+      "learning_rate": 9.900319999999999e-07,
+      "loss": 0.8958,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0255,
+      "grad_norm": 36.27665328979492,
+      "learning_rate": 9.89832e-07,
+      "loss": 0.7135,
+      "step": 2550
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 76.19625854492188,
+      "learning_rate": 9.896319999999998e-07,
+      "loss": 0.9015,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0265,
+      "grad_norm": 54.22828674316406,
+      "learning_rate": 9.894319999999998e-07,
+      "loss": 0.7887,
+      "step": 2650
+    },
+    {
+      "epoch": 0.027,
+      "grad_norm": 77.00263214111328,
+      "learning_rate": 9.89232e-07,
+      "loss": 0.6704,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 65.53499603271484,
+      "learning_rate": 9.89032e-07,
+      "loss": 0.6049,
+      "step": 2750
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 97.3672103881836,
+      "learning_rate": 9.88832e-07,
+      "loss": 0.7043,
+      "step": 2800
+    },
+    {
+      "epoch": 0.0285,
+      "grad_norm": 80.95391082763672,
+      "learning_rate": 9.88632e-07,
+      "loss": 0.7861,
+      "step": 2850
+    },
+    {
+      "epoch": 0.029,
+      "grad_norm": 39.608619689941406,
+      "learning_rate": 9.88432e-07,
+      "loss": 0.7095,
+      "step": 2900
+    },
+    {
+      "epoch": 0.0295,
+      "grad_norm": 71.16398620605469,
+      "learning_rate": 9.88232e-07,
+      "loss": 0.5963,
+      "step": 2950
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 66.53075408935547,
+      "learning_rate": 9.88032e-07,
+      "loss": 0.7837,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0305,
+      "grad_norm": 77.58058166503906,
+      "learning_rate": 9.87832e-07,
+      "loss": 0.627,
+      "step": 3050
+    },
+    {
+      "epoch": 0.031,
+      "grad_norm": 60.44248962402344,
+      "learning_rate": 9.87632e-07,
+      "loss": 0.7445,
+      "step": 3100
+    },
+    {
+      "epoch": 0.0315,
+      "grad_norm": 69.0057144165039,
+      "learning_rate": 9.874319999999999e-07,
+      "loss": 0.9611,
+      "step": 3150
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 103.23316192626953,
+      "learning_rate": 9.872319999999999e-07,
+      "loss": 0.6921,
+      "step": 3200
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 63.068687438964844,
+      "learning_rate": 9.87032e-07,
+      "loss": 0.7081,
+      "step": 3250
+    },
+    {
+      "epoch": 0.033,
+      "grad_norm": 61.475379943847656,
+      "learning_rate": 9.86832e-07,
+      "loss": 0.8122,
+      "step": 3300
+    },
+    {
+      "epoch": 0.0335,
+      "grad_norm": 65.56523895263672,
+      "learning_rate": 9.866319999999998e-07,
+      "loss": 0.7693,
+      "step": 3350
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 62.29042434692383,
+      "learning_rate": 9.86432e-07,
+      "loss": 0.8332,
+      "step": 3400
+    },
+    {
+      "epoch": 0.0345,
+      "grad_norm": 66.55669403076172,
+      "learning_rate": 9.86232e-07,
+      "loss": 0.7765,
+      "step": 3450
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 41.560089111328125,
+      "learning_rate": 9.86032e-07,
+      "loss": 0.8187,
+      "step": 3500
+    },
+    {
+      "epoch": 0.0355,
+      "grad_norm": 59.37902069091797,
+      "learning_rate": 9.85832e-07,
+      "loss": 0.6688,
+      "step": 3550
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 24.928035736083984,
+      "learning_rate": 9.85632e-07,
+      "loss": 0.6727,
+      "step": 3600
+    },
+    {
+      "epoch": 0.0365,
+      "grad_norm": 10.977078437805176,
+      "learning_rate": 9.85432e-07,
+      "loss": 0.7937,
+      "step": 3650
+    },
+    {
+      "epoch": 0.037,
+      "grad_norm": 52.790130615234375,
+      "learning_rate": 9.85232e-07,
+      "loss": 0.9761,
+      "step": 3700
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 10.147725105285645,
+      "learning_rate": 9.85032e-07,
+      "loss": 0.703,
+      "step": 3750
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 17.01890754699707,
+      "learning_rate": 9.84832e-07,
+      "loss": 0.7169,
+      "step": 3800
+    },
+    {
+      "epoch": 0.0385,
+      "grad_norm": 9.722551345825195,
+      "learning_rate": 9.846319999999999e-07,
+      "loss": 0.802,
+      "step": 3850
+    },
+    {
+      "epoch": 0.039,
+      "grad_norm": 39.80437088012695,
+      "learning_rate": 9.844319999999999e-07,
+      "loss": 0.7751,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0395,
+      "grad_norm": 43.26958465576172,
+      "learning_rate": 9.842319999999999e-07,
+      "loss": 0.7675,
+      "step": 3950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 53.78465270996094,
+      "learning_rate": 9.84032e-07,
+      "loss": 0.683,
+      "step": 4000
+    },
+    {
+      "epoch": 0.0405,
+      "grad_norm": 7.1808319091796875,
+      "learning_rate": 9.83832e-07,
+      "loss": 0.7682,
+      "step": 4050
+    },
+    {
+      "epoch": 0.041,
+      "grad_norm": 75.9840316772461,
+      "learning_rate": 9.836319999999998e-07,
+      "loss": 0.6552,
+      "step": 4100
+    },
+    {
+      "epoch": 0.0415,
+      "grad_norm": 22.792072296142578,
+      "learning_rate": 9.83432e-07,
+      "loss": 0.7268,
+      "step": 4150
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 29.64811897277832,
+      "learning_rate": 9.83232e-07,
+      "loss": 0.6025,
+      "step": 4200
+    },
+    {
+      "epoch": 0.0425,
+      "grad_norm": 93.7188720703125,
+      "learning_rate": 9.83032e-07,
+      "loss": 0.7603,
+      "step": 4250
+    },
+    {
+      "epoch": 0.043,
+      "grad_norm": 81.36428833007812,
+      "learning_rate": 9.82832e-07,
+      "loss": 0.7969,
+      "step": 4300
+    },
+    {
+      "epoch": 0.0435,
+      "grad_norm": 72.37635040283203,
+      "learning_rate": 9.82632e-07,
+      "loss": 0.6322,
+      "step": 4350
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 58.41753387451172,
+      "learning_rate": 9.82432e-07,
+      "loss": 0.6965,
+      "step": 4400
+    },
+    {
+      "epoch": 0.0445,
+      "grad_norm": 26.843284606933594,
+      "learning_rate": 9.82232e-07,
+      "loss": 0.622,
+      "step": 4450
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 80.62271881103516,
+      "learning_rate": 9.82032e-07,
+      "loss": 0.7182,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0455,
+      "grad_norm": 62.724483489990234,
+      "learning_rate": 9.81832e-07,
+      "loss": 0.6782,
+      "step": 4550
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 39.624332427978516,
+      "learning_rate": 9.816319999999999e-07,
+      "loss": 0.5628,
+      "step": 4600
+    },
+    {
+      "epoch": 0.0465,
+      "grad_norm": 30.11197853088379,
+      "learning_rate": 9.814319999999999e-07,
+      "loss": 0.6594,
+      "step": 4650
+    },
+    {
+      "epoch": 0.047,
+      "grad_norm": 47.84878158569336,
+      "learning_rate": 9.812319999999998e-07,
+      "loss": 0.5869,
+      "step": 4700
+    },
+    {
+      "epoch": 0.0475,
+      "grad_norm": 31.320173263549805,
+      "learning_rate": 9.81032e-07,
+      "loss": 0.7272,
+      "step": 4750
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 57.74075698852539,
+      "learning_rate": 9.80832e-07,
+      "loss": 0.7878,
+      "step": 4800
+    },
+    {
+      "epoch": 0.0485,
+      "grad_norm": 66.6219711303711,
+      "learning_rate": 9.806319999999998e-07,
+      "loss": 0.7573,
+      "step": 4850
+    },
+    {
+      "epoch": 0.049,
+      "grad_norm": 28.821975708007812,
+      "learning_rate": 9.80432e-07,
+      "loss": 0.7238,
+      "step": 4900
+    },
+    {
+      "epoch": 0.0495,
+      "grad_norm": 70.03323364257812,
+      "learning_rate": 9.80232e-07,
+      "loss": 0.7167,
+      "step": 4950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 42.992374420166016,
+      "learning_rate": 9.80032e-07,
+      "loss": 0.6073,
+      "step": 5000
+    },
+    {
+      "epoch": 0.0505,
+      "grad_norm": 80.54195404052734,
+      "learning_rate": 9.79832e-07,
+      "loss": 0.7736,
+      "step": 5050
+    },
+    {
+      "epoch": 0.051,
+      "grad_norm": 58.23442459106445,
+      "learning_rate": 9.79632e-07,
+      "loss": 0.7848,
+      "step": 5100
+    },
+    {
+      "epoch": 0.0515,
+      "grad_norm": 51.8128662109375,
+      "learning_rate": 9.79432e-07,
+      "loss": 0.7043,
+      "step": 5150
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 68.98834228515625,
+      "learning_rate": 9.79232e-07,
+      "loss": 0.7293,
+      "step": 5200
+    },
+    {
+      "epoch": 0.0525,
+      "grad_norm": 75.1903305053711,
+      "learning_rate": 9.79032e-07,
+      "loss": 0.7225,
+      "step": 5250
+    },
+    {
+      "epoch": 0.053,
+      "grad_norm": 46.63593292236328,
+      "learning_rate": 9.78832e-07,
+      "loss": 0.7507,
+      "step": 5300
+    },
+    {
+      "epoch": 0.0535,
+      "grad_norm": 54.36689376831055,
+      "learning_rate": 9.786319999999999e-07,
+      "loss": 0.8457,
+      "step": 5350
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 20.45354461669922,
+      "learning_rate": 9.784319999999999e-07,
+      "loss": 0.621,
+      "step": 5400
+    },
+    {
+      "epoch": 0.0545,
+      "grad_norm": 49.178565979003906,
+      "learning_rate": 9.78232e-07,
+      "loss": 0.8933,
+      "step": 5450
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 1.9964510202407837,
+      "learning_rate": 9.78032e-07,
+      "loss": 0.5691,
+      "step": 5500
+    },
+    {
+      "epoch": 0.0555,
+      "grad_norm": 74.9593734741211,
+      "learning_rate": 9.77832e-07,
+      "loss": 0.614,
+      "step": 5550
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 62.81483459472656,
+      "learning_rate": 9.77632e-07,
+      "loss": 0.8412,
+      "step": 5600
+    },
+    {
+      "epoch": 0.0565,
+      "grad_norm": 0.782301664352417,
+      "learning_rate": 9.77432e-07,
+      "loss": 0.6128,
+      "step": 5650
+    },
+    {
+      "epoch": 0.057,
+      "grad_norm": 62.49407196044922,
+      "learning_rate": 9.77232e-07,
+      "loss": 0.6208,
+      "step": 5700
+    },
+    {
+      "epoch": 0.0575,
+      "grad_norm": 39.935970306396484,
+      "learning_rate": 9.77032e-07,
+      "loss": 0.5399,
+      "step": 5750
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 25.412355422973633,
+      "learning_rate": 9.76832e-07,
+      "loss": 0.6784,
+      "step": 5800
+    },
+    {
+      "epoch": 0.0585,
+      "grad_norm": 64.27447509765625,
+      "learning_rate": 9.76632e-07,
+      "loss": 0.7829,
+      "step": 5850
+    },
+    {
+      "epoch": 0.059,
+      "grad_norm": 85.29767608642578,
+      "learning_rate": 9.76432e-07,
+      "loss": 0.649,
+      "step": 5900
+    },
+    {
+      "epoch": 0.0595,
+      "grad_norm": 36.06076431274414,
+      "learning_rate": 9.76232e-07,
+      "loss": 0.7838,
+      "step": 5950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 84.43746948242188,
+      "learning_rate": 9.760319999999999e-07,
+      "loss": 0.6822,
+      "step": 6000
+    },
+    {
+      "epoch": 0.0605,
+      "grad_norm": 74.78901672363281,
+      "learning_rate": 9.75832e-07,
+      "loss": 0.6003,
+      "step": 6050
+    },
+    {
+      "epoch": 0.061,
+      "grad_norm": 70.3706283569336,
+      "learning_rate": 9.75632e-07,
+      "loss": 0.6829,
+      "step": 6100
+    },
+    {
+      "epoch": 0.0615,
+      "grad_norm": 47.27094650268555,
+      "learning_rate": 9.754319999999998e-07,
+      "loss": 0.6415,
+      "step": 6150
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 19.190216064453125,
+      "learning_rate": 9.75232e-07,
+      "loss": 0.5789,
+      "step": 6200
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 82.82536315917969,
+      "learning_rate": 9.75032e-07,
+      "loss": 0.6222,
+      "step": 6250
+    },
+    {
+      "epoch": 0.063,
+      "grad_norm": 72.95569610595703,
+      "learning_rate": 9.74832e-07,
+      "loss": 0.6288,
+      "step": 6300
+    },
+    {
+      "epoch": 0.0635,
+      "grad_norm": 19.769786834716797,
+      "learning_rate": 9.74632e-07,
+      "loss": 0.7223,
+      "step": 6350
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 22.352313995361328,
+      "learning_rate": 9.74432e-07,
+      "loss": 0.6113,
+      "step": 6400
+    },
+    {
+      "epoch": 0.0645,
+      "grad_norm": 42.68896484375,
+      "learning_rate": 9.74232e-07,
+      "loss": 0.5704,
+      "step": 6450
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 8.79783821105957,
+      "learning_rate": 9.74032e-07,
+      "loss": 0.7296,
+      "step": 6500
+    },
+    {
+      "epoch": 0.0655,
+      "grad_norm": 19.08173942565918,
+      "learning_rate": 9.73832e-07,
+      "loss": 0.5812,
+      "step": 6550
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 7.165023326873779,
+      "learning_rate": 9.73632e-07,
+      "loss": 0.7078,
+      "step": 6600
+    },
+    {
+      "epoch": 0.0665,
+      "grad_norm": 58.18810272216797,
+      "learning_rate": 9.73432e-07,
+      "loss": 0.6484,
+      "step": 6650
+    },
+    {
+      "epoch": 0.067,
+      "grad_norm": 50.902000427246094,
+      "learning_rate": 9.732319999999999e-07,
+      "loss": 0.7417,
+      "step": 6700
+    },
+    {
+      "epoch": 0.0675,
+      "grad_norm": 15.194981575012207,
+      "learning_rate": 9.730319999999999e-07,
+      "loss": 0.66,
+      "step": 6750
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 12.29551887512207,
+      "learning_rate": 9.72832e-07,
+      "loss": 0.6219,
+      "step": 6800
+    },
+    {
+      "epoch": 0.0685,
+      "grad_norm": 33.02836608886719,
+      "learning_rate": 9.72632e-07,
+      "loss": 0.7864,
+      "step": 6850
+    },
+    {
+      "epoch": 0.069,
+      "grad_norm": 61.996707916259766,
+      "learning_rate": 9.724319999999998e-07,
+      "loss": 0.7458,
+      "step": 6900
+    },
+    {
+      "epoch": 0.0695,
+      "grad_norm": 50.779640197753906,
+      "learning_rate": 9.72232e-07,
+      "loss": 0.6969,
+      "step": 6950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 63.49092483520508,
+      "learning_rate": 9.72032e-07,
+      "loss": 0.6315,
+      "step": 7000
+    },
+    {
+      "epoch": 0.0705,
+      "grad_norm": 2.9683051109313965,
+      "learning_rate": 9.71832e-07,
+      "loss": 0.6708,
+      "step": 7050
+    },
+    {
+      "epoch": 0.071,
+      "grad_norm": 63.99093246459961,
+      "learning_rate": 9.71632e-07,
+      "loss": 0.5949,
+      "step": 7100
+    },
+    {
+      "epoch": 0.0715,
+      "grad_norm": 59.804561614990234,
+      "learning_rate": 9.71432e-07,
+      "loss": 0.6125,
+      "step": 7150
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 52.34803009033203,
+      "learning_rate": 9.71232e-07,
+      "loss": 0.6553,
+      "step": 7200
+    },
+    {
+      "epoch": 0.0725,
+      "grad_norm": 48.697906494140625,
+      "learning_rate": 9.71032e-07,
+      "loss": 0.7065,
+      "step": 7250
+    },
+    {
+      "epoch": 0.073,
+      "grad_norm": 22.473642349243164,
+      "learning_rate": 9.70832e-07,
+      "loss": 0.5137,
+      "step": 7300
+    },
+    {
+      "epoch": 0.0735,
+      "grad_norm": 108.86209106445312,
+      "learning_rate": 9.706320000000001e-07,
+      "loss": 0.5885,
+      "step": 7350
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 97.51536560058594,
+      "learning_rate": 9.704319999999999e-07,
+      "loss": 0.725,
+      "step": 7400
+    },
+    {
+      "epoch": 0.0745,
+      "grad_norm": 90.5220947265625,
+      "learning_rate": 9.702319999999999e-07,
+      "loss": 0.6542,
+      "step": 7450
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 79.1460189819336,
+      "learning_rate": 9.70032e-07,
+      "loss": 0.7872,
+      "step": 7500
+    },
+    {
+      "epoch": 0.0755,
+      "grad_norm": 11.08342170715332,
+      "learning_rate": 9.69832e-07,
+      "loss": 0.6952,
+      "step": 7550
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 70.80834197998047,
+      "learning_rate": 9.69632e-07,
+      "loss": 0.7243,
+      "step": 7600
+    },
+    {
+      "epoch": 0.0765,
+      "grad_norm": 51.24228286743164,
+      "learning_rate": 9.69432e-07,
+      "loss": 0.6433,
+      "step": 7650
+    },
+    {
+      "epoch": 0.077,
+      "grad_norm": 65.85570526123047,
+      "learning_rate": 9.69232e-07,
+      "loss": 0.7118,
+      "step": 7700
+    },
+    {
+      "epoch": 0.0775,
+      "grad_norm": 77.98046112060547,
+      "learning_rate": 9.69032e-07,
+      "loss": 0.6795,
+      "step": 7750
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 58.70011901855469,
+      "learning_rate": 9.68832e-07,
+      "loss": 0.677,
+      "step": 7800
+    },
+    {
+      "epoch": 0.0785,
+      "grad_norm": 47.72407150268555,
+      "learning_rate": 9.68632e-07,
+      "loss": 0.8678,
+      "step": 7850
+    },
+    {
+      "epoch": 0.079,
+      "grad_norm": 77.00784301757812,
+      "learning_rate": 9.684359999999998e-07,
+      "loss": 0.61,
+      "step": 7900
+    },
+    {
+      "epoch": 0.0795,
+      "grad_norm": 82.65626525878906,
+      "learning_rate": 9.68236e-07,
+      "loss": 0.6014,
+      "step": 7950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 64.11064147949219,
+      "learning_rate": 9.68036e-07,
+      "loss": 0.5775,
+      "step": 8000
+    },
+    {
+      "epoch": 0.0805,
+      "grad_norm": 75.26988220214844,
+      "learning_rate": 9.67836e-07,
+      "loss": 0.7189,
+      "step": 8050
+    },
+    {
+      "epoch": 0.081,
+      "grad_norm": 53.913936614990234,
+      "learning_rate": 9.67636e-07,
+      "loss": 0.6397,
+      "step": 8100
+    },
+    {
+      "epoch": 0.0815,
+      "grad_norm": 22.316036224365234,
+      "learning_rate": 9.67436e-07,
+      "loss": 0.5865,
+      "step": 8150
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 25.347740173339844,
+      "learning_rate": 9.67236e-07,
+      "loss": 0.6304,
+      "step": 8200
+    },
+    {
+      "epoch": 0.0825,
+      "grad_norm": 59.95059585571289,
+      "learning_rate": 9.67036e-07,
+      "loss": 0.5163,
+      "step": 8250
+    },
+    {
+      "epoch": 0.083,
+      "grad_norm": 35.91787338256836,
+      "learning_rate": 9.66836e-07,
+      "loss": 0.6994,
+      "step": 8300
+    },
+    {
+      "epoch": 0.0835,
+      "grad_norm": 10.109200477600098,
+      "learning_rate": 9.66636e-07,
+      "loss": 0.6286,
+      "step": 8350
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 25.906421661376953,
+      "learning_rate": 9.66436e-07,
+      "loss": 0.6338,
+      "step": 8400
+    },
+    {
+      "epoch": 0.0845,
+      "grad_norm": 79.9737777709961,
+      "learning_rate": 9.662359999999999e-07,
+      "loss": 0.8442,
+      "step": 8450
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 41.02357482910156,
+      "learning_rate": 9.660359999999999e-07,
+      "loss": 0.7414,
+      "step": 8500
+    },
+    {
+      "epoch": 0.0855,
+      "grad_norm": 44.978729248046875,
+      "learning_rate": 9.65836e-07,
+      "loss": 0.8366,
+      "step": 8550
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 55.472808837890625,
+      "learning_rate": 9.65636e-07,
+      "loss": 0.5963,
+      "step": 8600
+    },
+    {
+      "epoch": 0.0865,
+      "grad_norm": 11.16412353515625,
+      "learning_rate": 9.654359999999998e-07,
+      "loss": 0.6936,
+      "step": 8650
+    },
+    {
+      "epoch": 0.087,
+      "grad_norm": 83.75244903564453,
+      "learning_rate": 9.65236e-07,
+      "loss": 0.6425,
+      "step": 8700
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 54.668190002441406,
+      "learning_rate": 9.65036e-07,
+      "loss": 0.6609,
+      "step": 8750
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 78.21800231933594,
+      "learning_rate": 9.64836e-07,
+      "loss": 0.6267,
+      "step": 8800
+    },
+    {
+      "epoch": 0.0885,
+      "grad_norm": 13.47957706451416,
+      "learning_rate": 9.64636e-07,
+      "loss": 0.689,
+      "step": 8850
+    },
+    {
+      "epoch": 0.089,
+      "grad_norm": 15.554354667663574,
+      "learning_rate": 9.64436e-07,
+      "loss": 0.7207,
+      "step": 8900
+    },
+    {
+      "epoch": 0.0895,
+      "grad_norm": 77.30255126953125,
+      "learning_rate": 9.64236e-07,
+      "loss": 0.5907,
+      "step": 8950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 50.847564697265625,
+      "learning_rate": 9.64036e-07,
+      "loss": 0.623,
+      "step": 9000
+    },
+    {
+      "epoch": 0.0905,
+      "grad_norm": 20.90938377380371,
+      "learning_rate": 9.63836e-07,
+      "loss": 0.7254,
+      "step": 9050
+    },
+    {
+      "epoch": 0.091,
+      "grad_norm": 77.27519989013672,
+      "learning_rate": 9.63636e-07,
+      "loss": 0.6997,
+      "step": 9100
+    },
+    {
+      "epoch": 0.0915,
+      "grad_norm": 71.25733947753906,
+      "learning_rate": 9.634359999999999e-07,
+      "loss": 0.6436,
+      "step": 9150
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 47.3591423034668,
+      "learning_rate": 9.632359999999999e-07,
+      "loss": 0.6289,
+      "step": 9200
+    },
+    {
+      "epoch": 0.0925,
+      "grad_norm": 0.31778019666671753,
+      "learning_rate": 9.630359999999999e-07,
+      "loss": 0.58,
+      "step": 9250
+    },
+    {
+      "epoch": 0.093,
+      "grad_norm": 73.01952362060547,
+      "learning_rate": 9.62836e-07,
+      "loss": 0.6876,
+      "step": 9300
+    },
+    {
+      "epoch": 0.0935,
+      "grad_norm": 79.24958038330078,
+      "learning_rate": 9.62636e-07,
+      "loss": 0.7076,
+      "step": 9350
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 33.39700698852539,
+      "learning_rate": 9.624359999999998e-07,
+      "loss": 0.6385,
+      "step": 9400
+    },
+    {
+      "epoch": 0.0945,
+      "grad_norm": 6.0091376304626465,
+      "learning_rate": 9.62236e-07,
+      "loss": 0.6001,
+      "step": 9450
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 93.4347152709961,
+      "learning_rate": 9.62036e-07,
+      "loss": 0.5681,
+      "step": 9500
+    },
+    {
+      "epoch": 0.0955,
+      "grad_norm": 124.39002227783203,
+      "learning_rate": 9.61836e-07,
+      "loss": 0.5942,
+      "step": 9550
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 26.654417037963867,
+      "learning_rate": 9.61636e-07,
+      "loss": 0.5937,
+      "step": 9600
+    },
+    {
+      "epoch": 0.0965,
+      "grad_norm": 84.6335678100586,
+      "learning_rate": 9.61436e-07,
+      "loss": 0.6569,
+      "step": 9650
+    },
+    {
+      "epoch": 0.097,
+      "grad_norm": 68.39822387695312,
+      "learning_rate": 9.61236e-07,
+      "loss": 0.6662,
+      "step": 9700
+    },
+    {
+      "epoch": 0.0975,
+      "grad_norm": 46.38802719116211,
+      "learning_rate": 9.61036e-07,
+      "loss": 0.6765,
+      "step": 9750
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 12.477609634399414,
+      "learning_rate": 9.60836e-07,
+      "loss": 0.6491,
+      "step": 9800
+    },
+    {
+      "epoch": 0.0985,
+      "grad_norm": 7.226109027862549,
+      "learning_rate": 9.606360000000001e-07,
+      "loss": 0.6194,
+      "step": 9850
+    },
+    {
+      "epoch": 0.099,
+      "grad_norm": 47.4417724609375,
+      "learning_rate": 9.604359999999999e-07,
+      "loss": 0.7301,
+      "step": 9900
+    },
+    {
+      "epoch": 0.0995,
+      "grad_norm": 91.45806884765625,
+      "learning_rate": 9.602359999999999e-07,
+      "loss": 0.6963,
+      "step": 9950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 40.271846771240234,
+      "learning_rate": 9.60036e-07,
+      "loss": 0.7493,
+      "step": 10000
+    },
+    {
+      "epoch": 0.1005,
+      "grad_norm": 87.0700912475586,
+      "learning_rate": 9.59836e-07,
+      "loss": 0.558,
+      "step": 10050
+    },
+    {
+      "epoch": 0.101,
+      "grad_norm": 89.44054412841797,
+      "learning_rate": 9.5964e-07,
+      "loss": 0.4637,
+      "step": 10100
+    },
+    {
+      "epoch": 0.1015,
+      "grad_norm": 104.4756088256836,
+      "learning_rate": 9.5944e-07,
+      "loss": 0.6641,
+      "step": 10150
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 77.9275131225586,
+      "learning_rate": 9.5924e-07,
+      "loss": 0.4791,
+      "step": 10200
+    },
+    {
+      "epoch": 0.1025,
+      "grad_norm": 113.33071899414062,
+      "learning_rate": 9.590399999999999e-07,
+      "loss": 0.5757,
+      "step": 10250
+    },
+    {
+      "epoch": 0.103,
+      "grad_norm": 74.14317321777344,
+      "learning_rate": 9.5884e-07,
+      "loss": 0.7051,
+      "step": 10300
+    },
+    {
+      "epoch": 0.1035,
+      "grad_norm": 91.95326232910156,
+      "learning_rate": 9.5864e-07,
+      "loss": 0.7315,
+      "step": 10350
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 72.8747329711914,
+      "learning_rate": 9.584399999999998e-07,
+      "loss": 0.7391,
+      "step": 10400
+    },
+    {
+      "epoch": 0.1045,
+      "grad_norm": 61.38804244995117,
+      "learning_rate": 9.5824e-07,
+      "loss": 0.6073,
+      "step": 10450
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 65.2311782836914,
+      "learning_rate": 9.5804e-07,
+      "loss": 0.792,
+      "step": 10500
+    },
+    {
+      "epoch": 0.1055,
+      "grad_norm": 52.02727508544922,
+      "learning_rate": 9.5784e-07,
+      "loss": 0.5261,
+      "step": 10550
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 34.90068054199219,
+      "learning_rate": 9.5764e-07,
+      "loss": 0.5306,
+      "step": 10600
+    },
+    {
+      "epoch": 0.1065,
+      "grad_norm": 58.16232681274414,
+      "learning_rate": 9.5744e-07,
+      "loss": 0.7673,
+      "step": 10650
+    },
+    {
+      "epoch": 0.107,
+      "grad_norm": 15.396655082702637,
+      "learning_rate": 9.5724e-07,
+      "loss": 0.5671,
+      "step": 10700
+    },
+    {
+      "epoch": 0.1075,
+      "grad_norm": 15.548702239990234,
+      "learning_rate": 9.5704e-07,
+      "loss": 0.7041,
+      "step": 10750
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 101.34947204589844,
+      "learning_rate": 9.5684e-07,
+      "loss": 0.7207,
+      "step": 10800
+    },
+    {
+      "epoch": 0.1085,
+      "grad_norm": 6.5170207023620605,
+      "learning_rate": 9.5664e-07,
+      "loss": 0.5903,
+      "step": 10850
+    },
+    {
+      "epoch": 0.109,
+      "grad_norm": 36.6441764831543,
+      "learning_rate": 9.5644e-07,
+      "loss": 0.6876,
+      "step": 10900
+    },
+    {
+      "epoch": 0.1095,
+      "grad_norm": 63.76533508300781,
+      "learning_rate": 9.562399999999999e-07,
+      "loss": 0.6007,
+      "step": 10950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 52.73303985595703,
+      "learning_rate": 9.560399999999999e-07,
+      "loss": 0.6017,
+      "step": 11000
+    },
+    {
+      "epoch": 0.1105,
+      "grad_norm": 62.94703674316406,
+      "learning_rate": 9.5584e-07,
+      "loss": 0.6171,
+      "step": 11050
+    },
+    {
+      "epoch": 0.111,
+      "grad_norm": 60.48774337768555,
+      "learning_rate": 9.5564e-07,
+      "loss": 0.5815,
+      "step": 11100
+    },
+    {
+      "epoch": 0.1115,
+      "grad_norm": 84.18730163574219,
+      "learning_rate": 9.554399999999998e-07,
+      "loss": 0.6942,
+      "step": 11150
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 81.33743286132812,
+      "learning_rate": 9.5524e-07,
+      "loss": 0.5393,
+      "step": 11200
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 58.714378356933594,
+      "learning_rate": 9.5504e-07,
+      "loss": 0.532,
+      "step": 11250
+    },
+    {
+      "epoch": 0.113,
+      "grad_norm": 97.91901397705078,
+      "learning_rate": 9.5484e-07,
+      "loss": 0.6323,
+      "step": 11300
+    },
+    {
+      "epoch": 0.1135,
+      "grad_norm": 5.727772235870361,
+      "learning_rate": 9.5464e-07,
+      "loss": 0.5041,
+      "step": 11350
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 39.63825988769531,
+      "learning_rate": 9.5444e-07,
+      "loss": 0.5549,
+      "step": 11400
+    },
+    {
+      "epoch": 0.1145,
+      "grad_norm": 103.17711639404297,
+      "learning_rate": 9.5424e-07,
+      "loss": 0.5167,
+      "step": 11450
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 48.15481948852539,
+      "learning_rate": 9.5404e-07,
+      "loss": 0.7287,
+      "step": 11500
+    },
+    {
+      "epoch": 0.1155,
+      "grad_norm": 4.032690525054932,
+      "learning_rate": 9.5384e-07,
+      "loss": 0.64,
+      "step": 11550
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 22.449026107788086,
+      "learning_rate": 9.5364e-07,
+      "loss": 0.6243,
+      "step": 11600
+    },
+    {
+      "epoch": 0.1165,
+      "grad_norm": 76.77507019042969,
+      "learning_rate": 9.534399999999999e-07,
+      "loss": 0.6222,
+      "step": 11650
+    },
+    {
+      "epoch": 0.117,
+      "grad_norm": 3.4103989601135254,
+      "learning_rate": 9.5324e-07,
+      "loss": 0.7213,
+      "step": 11700
+    },
+    {
+      "epoch": 0.1175,
+      "grad_norm": 40.05355453491211,
+      "learning_rate": 9.5304e-07,
+      "loss": 0.6782,
+      "step": 11750
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 13.29817008972168,
+      "learning_rate": 9.5284e-07,
+      "loss": 0.7741,
+      "step": 11800
+    },
+    {
+      "epoch": 0.1185,
+      "grad_norm": 55.80624771118164,
+      "learning_rate": 9.5264e-07,
+      "loss": 0.7785,
+      "step": 11850
+    },
+    {
+      "epoch": 0.119,
+      "grad_norm": 75.11266326904297,
+      "learning_rate": 9.524399999999999e-07,
+      "loss": 0.7455,
+      "step": 11900
+    },
+    {
+      "epoch": 0.1195,
+      "grad_norm": 28.10260009765625,
+      "learning_rate": 9.522399999999999e-07,
+      "loss": 0.6025,
+      "step": 11950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 71.65565490722656,
+      "learning_rate": 9.5204e-07,
+      "loss": 0.7934,
+      "step": 12000
+    },
+    {
+      "epoch": 0.1205,
+      "grad_norm": 54.734107971191406,
+      "learning_rate": 9.5184e-07,
+      "loss": 0.6244,
+      "step": 12050
+    },
+    {
+      "epoch": 0.121,
+      "grad_norm": 3.3738136291503906,
+      "learning_rate": 9.5164e-07,
+      "loss": 0.51,
+      "step": 12100
+    },
+    {
+      "epoch": 0.1215,
+      "grad_norm": 50.44538116455078,
+      "learning_rate": 9.5144e-07,
+      "loss": 0.5917,
+      "step": 12150
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 22.168771743774414,
+      "learning_rate": 9.512399999999999e-07,
+      "loss": 0.5424,
+      "step": 12200
+    },
+    {
+      "epoch": 0.1225,
+      "grad_norm": 55.60480499267578,
+      "learning_rate": 9.510399999999999e-07,
+      "loss": 0.6332,
+      "step": 12250
+    },
+    {
+      "epoch": 0.123,
+      "grad_norm": 42.9905891418457,
+      "learning_rate": 9.5084e-07,
+      "loss": 0.5598,
+      "step": 12300
+    },
+    {
+      "epoch": 0.1235,
+      "grad_norm": 2.6760189533233643,
+      "learning_rate": 9.5064e-07,
+      "loss": 0.671,
+      "step": 12350
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 62.84537124633789,
+      "learning_rate": 9.504399999999999e-07,
+      "loss": 0.6672,
+      "step": 12400
+    },
+    {
+      "epoch": 0.1245,
+      "grad_norm": 60.886653900146484,
+      "learning_rate": 9.5024e-07,
+      "loss": 0.7067,
+      "step": 12450
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 84.8931655883789,
+      "learning_rate": 9.5004e-07,
+      "loss": 0.5629,
+      "step": 12500
+    },
+    {
+      "epoch": 0.1255,
+      "grad_norm": 0.7383530735969543,
+      "learning_rate": 9.498399999999999e-07,
+      "loss": 0.5731,
+      "step": 12550
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 146.97189331054688,
+      "learning_rate": 9.4964e-07,
+      "loss": 0.5422,
+      "step": 12600
+    },
+    {
+      "epoch": 0.1265,
+      "grad_norm": 13.385457038879395,
+      "learning_rate": 9.494399999999999e-07,
+      "loss": 0.6314,
+      "step": 12650
+    },
+    {
+      "epoch": 0.127,
+      "grad_norm": 9.101408004760742,
+      "learning_rate": 9.492399999999999e-07,
+      "loss": 0.6082,
+      "step": 12700
+    },
+    {
+      "epoch": 0.1275,
+      "grad_norm": 9.814948081970215,
+      "learning_rate": 9.4904e-07,
+      "loss": 0.6427,
+      "step": 12750
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 1.8460052013397217,
+      "learning_rate": 9.4884e-07,
+      "loss": 0.6449,
+      "step": 12800
+    },
+    {
+      "epoch": 0.1285,
+      "grad_norm": 2.2399051189422607,
+      "learning_rate": 9.4864e-07,
+      "loss": 0.6494,
+      "step": 12850
+    },
+    {
+      "epoch": 0.129,
+      "grad_norm": 51.1398811340332,
+      "learning_rate": 9.484399999999999e-07,
+      "loss": 0.5733,
+      "step": 12900
+    },
+    {
+      "epoch": 0.1295,
+      "grad_norm": 48.945919036865234,
+      "learning_rate": 9.482399999999999e-07,
+      "loss": 0.6088,
+      "step": 12950
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 55.05748748779297,
+      "learning_rate": 9.480399999999999e-07,
+      "loss": 0.5011,
+      "step": 13000
+    },
+    {
+      "epoch": 0.1305,
+      "grad_norm": 13.456110954284668,
+      "learning_rate": 9.4784e-07,
+      "loss": 0.7739,
+      "step": 13050
+    },
+    {
+      "epoch": 0.131,
+      "grad_norm": 76.10543060302734,
+      "learning_rate": 9.4764e-07,
+      "loss": 0.6818,
+      "step": 13100
+    },
+    {
+      "epoch": 0.1315,
+      "grad_norm": 51.6418571472168,
+      "learning_rate": 9.474439999999999e-07,
+      "loss": 0.525,
+      "step": 13150
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 17.212923049926758,
+      "learning_rate": 9.47244e-07,
+      "loss": 0.5746,
+      "step": 13200
+    },
+    {
+      "epoch": 0.1325,
+      "grad_norm": 53.69514465332031,
+      "learning_rate": 9.470439999999999e-07,
+      "loss": 0.5003,
+      "step": 13250
+    },
+    {
+      "epoch": 0.133,
+      "grad_norm": 40.141632080078125,
+      "learning_rate": 9.468439999999999e-07,
+      "loss": 0.6027,
+      "step": 13300
+    },
+    {
+      "epoch": 0.1335,
+      "grad_norm": 61.87038040161133,
+      "learning_rate": 9.46644e-07,
+      "loss": 0.6072,
+      "step": 13350
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 122.38882446289062,
+      "learning_rate": 9.464439999999999e-07,
+      "loss": 0.6451,
+      "step": 13400
+    },
+    {
+      "epoch": 0.1345,
+      "grad_norm": 56.08554458618164,
+      "learning_rate": 9.462439999999999e-07,
+      "loss": 0.6374,
+      "step": 13450
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 81.84549713134766,
+      "learning_rate": 9.46044e-07,
+      "loss": 0.5401,
+      "step": 13500
+    },
+    {
+      "epoch": 0.1355,
+      "grad_norm": 118.67034912109375,
+      "learning_rate": 9.45844e-07,
+      "loss": 0.6624,
+      "step": 13550
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 57.54740905761719,
+      "learning_rate": 9.45644e-07,
+      "loss": 0.5806,
+      "step": 13600
+    },
+    {
+      "epoch": 0.1365,
+      "grad_norm": 105.1045913696289,
+      "learning_rate": 9.454439999999999e-07,
+      "loss": 0.5867,
+      "step": 13650
+    },
+    {
+      "epoch": 0.137,
+      "grad_norm": 25.346324920654297,
+      "learning_rate": 9.452439999999999e-07,
+      "loss": 0.6926,
+      "step": 13700
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 91.3180160522461,
+      "learning_rate": 9.45044e-07,
+      "loss": 0.5217,
+      "step": 13750
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 79.1546630859375,
+      "learning_rate": 9.44844e-07,
+      "loss": 0.6172,
+      "step": 13800
+    },
+    {
+      "epoch": 0.1385,
+      "grad_norm": 92.9755630493164,
+      "learning_rate": 9.44644e-07,
+      "loss": 0.7169,
+      "step": 13850
+    },
+    {
+      "epoch": 0.139,
+      "grad_norm": 33.35745620727539,
+      "learning_rate": 9.44444e-07,
+      "loss": 0.6807,
+      "step": 13900
+    },
+    {
+      "epoch": 0.1395,
+      "grad_norm": 14.041089057922363,
+      "learning_rate": 9.442439999999999e-07,
+      "loss": 0.5662,
+      "step": 13950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.25873947143555,
+      "learning_rate": 9.440439999999999e-07,
+      "loss": 0.6316,
+      "step": 14000
+    },
+    {
+      "epoch": 0.1405,
+      "grad_norm": 59.24448013305664,
+      "learning_rate": 9.43844e-07,
+      "loss": 0.6294,
+      "step": 14050
+    },
+    {
+      "epoch": 0.141,
+      "grad_norm": 68.7812728881836,
+      "learning_rate": 9.43644e-07,
+      "loss": 0.6128,
+      "step": 14100
+    },
+    {
+      "epoch": 0.1415,
+      "grad_norm": 66.03839874267578,
+      "learning_rate": 9.434439999999999e-07,
+      "loss": 0.7409,
+      "step": 14150
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 28.09392738342285,
+      "learning_rate": 9.43244e-07,
+      "loss": 0.7238,
+      "step": 14200
+    },
+    {
+      "epoch": 0.1425,
+      "grad_norm": 11.177745819091797,
+      "learning_rate": 9.43044e-07,
+      "loss": 0.5512,
+      "step": 14250
+    },
+    {
+      "epoch": 0.143,
+      "grad_norm": 77.08145141601562,
+      "learning_rate": 9.428439999999999e-07,
+      "loss": 0.558,
+      "step": 14300
+    },
+    {
+      "epoch": 0.1435,
+      "grad_norm": 32.226314544677734,
+      "learning_rate": 9.42644e-07,
+      "loss": 0.6093,
+      "step": 14350
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 39.99115753173828,
+      "learning_rate": 9.424439999999999e-07,
+      "loss": 0.511,
+      "step": 14400
+    },
+    {
+      "epoch": 0.1445,
+      "grad_norm": 45.316959381103516,
+      "learning_rate": 9.422439999999999e-07,
+      "loss": 0.6866,
+      "step": 14450
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 9.914299011230469,
+      "learning_rate": 9.42044e-07,
+      "loss": 0.5704,
+      "step": 14500
+    },
+    {
+      "epoch": 0.1455,
+      "grad_norm": 112.46765899658203,
+      "learning_rate": 9.41844e-07,
+      "loss": 0.6679,
+      "step": 14550
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 18.95020866394043,
+      "learning_rate": 9.41644e-07,
+      "loss": 0.4952,
+      "step": 14600
+    },
+    {
+      "epoch": 0.1465,
+      "grad_norm": 74.44670867919922,
+      "learning_rate": 9.414439999999999e-07,
+      "loss": 0.5307,
+      "step": 14650
+    },
+    {
+      "epoch": 0.147,
+      "grad_norm": 1.2418839931488037,
+      "learning_rate": 9.412439999999999e-07,
+      "loss": 0.5571,
+      "step": 14700
+    },
+    {
+      "epoch": 0.1475,
+      "grad_norm": 49.43461608886719,
+      "learning_rate": 9.410439999999999e-07,
+      "loss": 0.4794,
+      "step": 14750
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 74.62863159179688,
+      "learning_rate": 9.40844e-07,
+      "loss": 0.5945,
+      "step": 14800
+    },
+    {
+      "epoch": 0.1485,
+      "grad_norm": 81.81336212158203,
+      "learning_rate": 9.40644e-07,
+      "loss": 0.62,
+      "step": 14850
+    },
+    {
+      "epoch": 0.149,
+      "grad_norm": 63.31863021850586,
+      "learning_rate": 9.404439999999999e-07,
+      "loss": 0.5821,
+      "step": 14900
+    },
+    {
+      "epoch": 0.1495,
+      "grad_norm": 42.47512435913086,
+      "learning_rate": 9.40244e-07,
+      "loss": 0.6294,
+      "step": 14950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 76.9931869506836,
+      "learning_rate": 9.400439999999999e-07,
+      "loss": 0.4261,
+      "step": 15000
+    },
+    {
+      "epoch": 0.1505,
+      "grad_norm": 61.6082763671875,
+      "learning_rate": 9.398439999999999e-07,
+      "loss": 0.671,
+      "step": 15050
+    },
+    {
+      "epoch": 0.151,
+      "grad_norm": 64.3666763305664,
+      "learning_rate": 9.39644e-07,
+      "loss": 0.7315,
+      "step": 15100
+    },
+    {
+      "epoch": 0.1515,
+      "grad_norm": 50.603111267089844,
+      "learning_rate": 9.394439999999999e-07,
+      "loss": 0.6032,
+      "step": 15150
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 88.94772338867188,
+      "learning_rate": 9.392439999999999e-07,
+      "loss": 0.5629,
+      "step": 15200
+    },
+    {
+      "epoch": 0.1525,
+      "grad_norm": 16.136049270629883,
+      "learning_rate": 9.39048e-07,
+      "loss": 0.547,
+      "step": 15250
+    },
+    {
+      "epoch": 0.153,
+      "grad_norm": 14.290252685546875,
+      "learning_rate": 9.38848e-07,
+      "loss": 0.5877,
+      "step": 15300
+    },
+    {
+      "epoch": 0.1535,
+      "grad_norm": 72.88977813720703,
+      "learning_rate": 9.38648e-07,
+      "loss": 0.6141,
+      "step": 15350
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 56.635902404785156,
+      "learning_rate": 9.384479999999999e-07,
+      "loss": 0.5268,
+      "step": 15400
+    },
+    {
+      "epoch": 0.1545,
+      "grad_norm": 36.33918762207031,
+      "learning_rate": 9.382479999999999e-07,
+      "loss": 0.6514,
+      "step": 15450
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 53.50493621826172,
+      "learning_rate": 9.380479999999999e-07,
+      "loss": 0.573,
+      "step": 15500
+    },
+    {
+      "epoch": 0.1555,
+      "grad_norm": 2.055042266845703,
+      "learning_rate": 9.37848e-07,
+      "loss": 0.6182,
+      "step": 15550
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 173.22987365722656,
+      "learning_rate": 9.37648e-07,
+      "loss": 0.6573,
+      "step": 15600
+    },
+    {
+      "epoch": 0.1565,
+      "grad_norm": 22.735443115234375,
+      "learning_rate": 9.374479999999999e-07,
+      "loss": 0.6504,
+      "step": 15650
+    },
+    {
+      "epoch": 0.157,
+      "grad_norm": 76.63526153564453,
+      "learning_rate": 9.37248e-07,
+      "loss": 0.6736,
+      "step": 15700
+    },
+    {
+      "epoch": 0.1575,
+      "grad_norm": 16.873645782470703,
+      "learning_rate": 9.370479999999999e-07,
+      "loss": 0.624,
+      "step": 15750
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 6.853653430938721,
+      "learning_rate": 9.368479999999999e-07,
+      "loss": 0.705,
+      "step": 15800
+    },
+    {
+      "epoch": 0.1585,
+      "grad_norm": 70.6560287475586,
+      "learning_rate": 9.36648e-07,
+      "loss": 0.603,
+      "step": 15850
+    },
+    {
+      "epoch": 0.159,
+      "grad_norm": 59.094749450683594,
+      "learning_rate": 9.364479999999999e-07,
+      "loss": 0.7677,
+      "step": 15900
+    },
+    {
+      "epoch": 0.1595,
+      "grad_norm": 7.74002742767334,
+      "learning_rate": 9.362479999999999e-07,
+      "loss": 0.5653,
+      "step": 15950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.42781639099121,
+      "learning_rate": 9.36048e-07,
+      "loss": 0.5747,
+      "step": 16000
+    },
+    {
+      "epoch": 0.1605,
+      "grad_norm": 56.63848876953125,
+      "learning_rate": 9.35848e-07,
+      "loss": 0.7402,
+      "step": 16050
+    },
+    {
+      "epoch": 0.161,
+      "grad_norm": 62.384342193603516,
+      "learning_rate": 9.35648e-07,
+      "loss": 0.6194,
+      "step": 16100
+    },
+    {
+      "epoch": 0.1615,
+      "grad_norm": 53.056766510009766,
+      "learning_rate": 9.354479999999999e-07,
+      "loss": 0.7444,
+      "step": 16150
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 82.36703491210938,
+      "learning_rate": 9.352479999999999e-07,
+      "loss": 0.5666,
+      "step": 16200
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 18.262683868408203,
+      "learning_rate": 9.35048e-07,
+      "loss": 0.5184,
+      "step": 16250
+    },
+    {
+      "epoch": 0.163,
+      "grad_norm": 96.85376739501953,
+      "learning_rate": 9.34848e-07,
+      "loss": 0.6913,
+      "step": 16300
+    },
+    {
+      "epoch": 0.1635,
+      "grad_norm": 35.92914962768555,
+      "learning_rate": 9.34648e-07,
+      "loss": 0.5972,
+      "step": 16350
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 33.37955856323242,
+      "learning_rate": 9.34448e-07,
+      "loss": 0.5924,
+      "step": 16400
+    },
+    {
+      "epoch": 0.1645,
+      "grad_norm": 33.09463882446289,
+      "learning_rate": 9.342479999999999e-07,
+      "loss": 0.5298,
+      "step": 16450
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 26.03669548034668,
+      "learning_rate": 9.340479999999999e-07,
+      "loss": 0.4642,
+      "step": 16500
+    },
+    {
+      "epoch": 0.1655,
+      "grad_norm": 20.21666717529297,
+      "learning_rate": 9.33848e-07,
+      "loss": 0.5442,
+      "step": 16550
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 23.451101303100586,
+      "learning_rate": 9.33648e-07,
+      "loss": 0.5649,
+      "step": 16600
+    },
+    {
+      "epoch": 0.1665,
+      "grad_norm": 8.985445022583008,
+      "learning_rate": 9.33448e-07,
+      "loss": 0.5812,
+      "step": 16650
+    },
+    {
+      "epoch": 0.167,
+      "grad_norm": 94.17129516601562,
+      "learning_rate": 9.33248e-07,
+      "loss": 0.5932,
+      "step": 16700
+    },
+    {
+      "epoch": 0.1675,
+      "grad_norm": 55.66262435913086,
+      "learning_rate": 9.33048e-07,
+      "loss": 0.5894,
+      "step": 16750
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 22.5528507232666,
+      "learning_rate": 9.328479999999999e-07,
+      "loss": 0.6531,
+      "step": 16800
+    },
+    {
+      "epoch": 0.1685,
+      "grad_norm": 1.4093486070632935,
+      "learning_rate": 9.32648e-07,
+      "loss": 0.6753,
+      "step": 16850
+    },
+    {
+      "epoch": 0.169,
+      "grad_norm": 2.8292222023010254,
+      "learning_rate": 9.32448e-07,
+      "loss": 0.6501,
+      "step": 16900
+    },
+    {
+      "epoch": 0.1695,
+      "grad_norm": 55.30830383300781,
+      "learning_rate": 9.322479999999999e-07,
+      "loss": 0.4652,
+      "step": 16950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9419615268707275,
+      "learning_rate": 9.32048e-07,
+      "loss": 0.5817,
+      "step": 17000
+    },
+    {
+      "epoch": 0.1705,
+      "grad_norm": 21.788589477539062,
+      "learning_rate": 9.31848e-07,
+      "loss": 0.4845,
+      "step": 17050
+    },
+    {
+      "epoch": 0.171,
+      "grad_norm": 51.12996292114258,
+      "learning_rate": 9.31648e-07,
+      "loss": 0.6086,
+      "step": 17100
+    },
+    {
+      "epoch": 0.1715,
+      "grad_norm": 29.451993942260742,
+      "learning_rate": 9.31448e-07,
+      "loss": 0.609,
+      "step": 17150
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 26.545246124267578,
+      "learning_rate": 9.312479999999999e-07,
+      "loss": 0.7538,
+      "step": 17200
+    },
+    {
+      "epoch": 0.1725,
+      "grad_norm": 88.69632720947266,
+      "learning_rate": 9.310479999999999e-07,
+      "loss": 0.629,
+      "step": 17250
+    },
+    {
+      "epoch": 0.173,
+      "grad_norm": 24.869253158569336,
+      "learning_rate": 9.30848e-07,
+      "loss": 0.6385,
+      "step": 17300
+    },
+    {
+      "epoch": 0.1735,
+      "grad_norm": 116.06053161621094,
+      "learning_rate": 9.30652e-07,
+      "loss": 0.6501,
+      "step": 17350
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 89.61216735839844,
+      "learning_rate": 9.304519999999999e-07,
+      "loss": 0.5625,
+      "step": 17400
+    },
+    {
+      "epoch": 0.1745,
+      "grad_norm": 0.683408260345459,
+      "learning_rate": 9.30252e-07,
+      "loss": 0.7442,
+      "step": 17450
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 43.17182540893555,
+      "learning_rate": 9.300519999999999e-07,
+      "loss": 0.7728,
+      "step": 17500
+    },
+    {
+      "epoch": 0.1755,
+      "grad_norm": 16.35731315612793,
+      "learning_rate": 9.298519999999999e-07,
+      "loss": 0.5745,
+      "step": 17550
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 8.577791213989258,
+      "learning_rate": 9.29652e-07,
+      "loss": 0.465,
+      "step": 17600
+    },
+    {
+      "epoch": 0.1765,
+      "grad_norm": 55.087440490722656,
+      "learning_rate": 9.294519999999999e-07,
+      "loss": 0.5468,
+      "step": 17650
+    },
+    {
+      "epoch": 0.177,
+      "grad_norm": 65.42967987060547,
+      "learning_rate": 9.292519999999999e-07,
+      "loss": 0.597,
+      "step": 17700
+    },
+    {
+      "epoch": 0.1775,
+      "grad_norm": 4.98286247253418,
+      "learning_rate": 9.29052e-07,
+      "loss": 0.5523,
+      "step": 17750
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 39.229949951171875,
+      "learning_rate": 9.28852e-07,
+      "loss": 0.6382,
+      "step": 17800
+    },
+    {
+      "epoch": 0.1785,
+      "grad_norm": 33.733394622802734,
+      "learning_rate": 9.286519999999999e-07,
+      "loss": 0.6871,
+      "step": 17850
+    },
+    {
+      "epoch": 0.179,
+      "grad_norm": 16.59604263305664,
+      "learning_rate": 9.28452e-07,
+      "loss": 0.6174,
+      "step": 17900
+    },
+    {
+      "epoch": 0.1795,
+      "grad_norm": 85.7757797241211,
+      "learning_rate": 9.282519999999999e-07,
+      "loss": 0.7438,
+      "step": 17950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 77.86770629882812,
+      "learning_rate": 9.280519999999999e-07,
+      "loss": 0.5513,
+      "step": 18000
+    },
+    {
+      "epoch": 0.1805,
+      "grad_norm": 4.501659393310547,
+      "learning_rate": 9.27852e-07,
+      "loss": 0.6059,
+      "step": 18050
+    },
+    {
+      "epoch": 0.181,
+      "grad_norm": 54.42782211303711,
+      "learning_rate": 9.27652e-07,
+      "loss": 0.4786,
+      "step": 18100
+    },
+    {
+      "epoch": 0.1815,
+      "grad_norm": 45.432151794433594,
+      "learning_rate": 9.274520000000001e-07,
+      "loss": 0.6646,
+      "step": 18150
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 111.56059265136719,
+      "learning_rate": 9.272519999999999e-07,
+      "loss": 0.6535,
+      "step": 18200
+    },
+    {
+      "epoch": 0.1825,
+      "grad_norm": 78.46669006347656,
+      "learning_rate": 9.270519999999999e-07,
+      "loss": 0.5738,
+      "step": 18250
+    },
+    {
+      "epoch": 0.183,
+      "grad_norm": 2.2137699127197266,
+      "learning_rate": 9.26852e-07,
+      "loss": 0.55,
+      "step": 18300
+    },
+    {
+      "epoch": 0.1835,
+      "grad_norm": 65.4900894165039,
+      "learning_rate": 9.26652e-07,
+      "loss": 0.6607,
+      "step": 18350
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 45.10904312133789,
+      "learning_rate": 9.26452e-07,
+      "loss": 0.6373,
+      "step": 18400
+    },
+    {
+      "epoch": 0.1845,
+      "grad_norm": 61.703250885009766,
+      "learning_rate": 9.26252e-07,
+      "loss": 0.4666,
+      "step": 18450
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 105.42646026611328,
+      "learning_rate": 9.26052e-07,
+      "loss": 0.6782,
+      "step": 18500
+    },
+    {
+      "epoch": 0.1855,
+      "grad_norm": 108.39800262451172,
+      "learning_rate": 9.258519999999999e-07,
+      "loss": 0.6013,
+      "step": 18550
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 48.339599609375,
+      "learning_rate": 9.25652e-07,
+      "loss": 0.5994,
+      "step": 18600
+    },
+    {
+      "epoch": 0.1865,
+      "grad_norm": 61.67754364013672,
+      "learning_rate": 9.25452e-07,
+      "loss": 0.5721,
+      "step": 18650
+    },
+    {
+      "epoch": 0.187,
+      "grad_norm": 33.096046447753906,
+      "learning_rate": 9.252519999999999e-07,
+      "loss": 0.6398,
+      "step": 18700
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 33.91469192504883,
+      "learning_rate": 9.25052e-07,
+      "loss": 0.4305,
+      "step": 18750
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 32.4777717590332,
+      "learning_rate": 9.24852e-07,
+      "loss": 0.6306,
+      "step": 18800
+    },
+    {
+      "epoch": 0.1885,
+      "grad_norm": 19.212507247924805,
+      "learning_rate": 9.24652e-07,
+      "loss": 0.6042,
+      "step": 18850
+    },
+    {
+      "epoch": 0.189,
+      "grad_norm": 193.7104949951172,
+      "learning_rate": 9.24452e-07,
+      "loss": 0.5036,
+      "step": 18900
+    },
+    {
+      "epoch": 0.1895,
+      "grad_norm": 77.8702392578125,
+      "learning_rate": 9.242519999999999e-07,
+      "loss": 0.6825,
+      "step": 18950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 53.27097702026367,
+      "learning_rate": 9.240519999999999e-07,
+      "loss": 0.6074,
+      "step": 19000
+    },
+    {
+      "epoch": 0.1905,
+      "grad_norm": 16.498302459716797,
+      "learning_rate": 9.23852e-07,
+      "loss": 0.497,
+      "step": 19050
+    },
+    {
+      "epoch": 0.191,
+      "grad_norm": 75.21617889404297,
+      "learning_rate": 9.23652e-07,
+      "loss": 0.6573,
+      "step": 19100
+    },
+    {
+      "epoch": 0.1915,
+      "grad_norm": 4.08469295501709,
+      "learning_rate": 9.23452e-07,
+      "loss": 0.5048,
+      "step": 19150
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 8.395776748657227,
+      "learning_rate": 9.23252e-07,
+      "loss": 0.5109,
+      "step": 19200
+    },
+    {
+      "epoch": 0.1925,
+      "grad_norm": 1.7910493612289429,
+      "learning_rate": 9.230519999999999e-07,
+      "loss": 0.5289,
+      "step": 19250
+    },
+    {
+      "epoch": 0.193,
+      "grad_norm": 67.36382293701172,
+      "learning_rate": 9.228519999999999e-07,
+      "loss": 0.7231,
+      "step": 19300
+    },
+    {
+      "epoch": 0.1935,
+      "grad_norm": 55.8679313659668,
+      "learning_rate": 9.22652e-07,
+      "loss": 0.5469,
+      "step": 19350
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 108.610107421875,
+      "learning_rate": 9.22452e-07,
+      "loss": 0.7487,
+      "step": 19400
+    },
+    {
+      "epoch": 0.1945,
+      "grad_norm": 60.358062744140625,
+      "learning_rate": 9.222519999999999e-07,
+      "loss": 0.5649,
+      "step": 19450
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 70.0567398071289,
+      "learning_rate": 9.22052e-07,
+      "loss": 0.526,
+      "step": 19500
+    },
+    {
+      "epoch": 0.1955,
+      "grad_norm": 23.887163162231445,
+      "learning_rate": 9.21856e-07,
+      "loss": 0.4628,
+      "step": 19550
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 15.568492889404297,
+      "learning_rate": 9.21656e-07,
+      "loss": 0.683,
+      "step": 19600
+    },
+    {
+      "epoch": 0.1965,
+      "grad_norm": 78.2328872680664,
+      "learning_rate": 9.21456e-07,
+      "loss": 0.6217,
+      "step": 19650
+    },
+    {
+      "epoch": 0.197,
+      "grad_norm": 16.458410263061523,
+      "learning_rate": 9.212559999999999e-07,
+      "loss": 0.6503,
+      "step": 19700
+    },
+    {
+      "epoch": 0.1975,
+      "grad_norm": 73.25173950195312,
+      "learning_rate": 9.210559999999999e-07,
+      "loss": 0.6609,
+      "step": 19750
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 47.52989959716797,
+      "learning_rate": 9.20856e-07,
+      "loss": 0.5237,
+      "step": 19800
+    },
+    {
+      "epoch": 0.1985,
+      "grad_norm": 62.07141876220703,
+      "learning_rate": 9.20656e-07,
+      "loss": 0.5814,
+      "step": 19850
+    },
+    {
+      "epoch": 0.199,
+      "grad_norm": 41.7489013671875,
+      "learning_rate": 9.20456e-07,
+      "loss": 0.6958,
+      "step": 19900
+    },
+    {
+      "epoch": 0.1995,
+      "grad_norm": 64.66690063476562,
+      "learning_rate": 9.20256e-07,
+      "loss": 0.5138,
+      "step": 19950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.058856010437012,
+      "learning_rate": 9.200559999999999e-07,
+      "loss": 0.5906,
+      "step": 20000
+    },
+    {
+      "epoch": 0.2005,
+      "grad_norm": 60.96815872192383,
+      "learning_rate": 9.198559999999999e-07,
+      "loss": 0.5804,
+      "step": 20050
+    },
+    {
+      "epoch": 0.201,
+      "grad_norm": 30.981765747070312,
+      "learning_rate": 9.19656e-07,
+      "loss": 0.8173,
+      "step": 20100
+    },
+    {
+      "epoch": 0.2015,
+      "grad_norm": 54.25685501098633,
+      "learning_rate": 9.19456e-07,
+      "loss": 0.5893,
+      "step": 20150
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 105.40068054199219,
+      "learning_rate": 9.192559999999999e-07,
+      "loss": 0.5352,
+      "step": 20200
+    },
+    {
+      "epoch": 0.2025,
+      "grad_norm": 36.076011657714844,
+      "learning_rate": 9.19056e-07,
+      "loss": 0.4749,
+      "step": 20250
+    },
+    {
+      "epoch": 0.203,
+      "grad_norm": 51.85771179199219,
+      "learning_rate": 9.18856e-07,
+      "loss": 0.6509,
+      "step": 20300
+    },
+    {
+      "epoch": 0.2035,
+      "grad_norm": 54.14777755737305,
+      "learning_rate": 9.186559999999999e-07,
+      "loss": 0.7679,
+      "step": 20350
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 90.21994018554688,
+      "learning_rate": 9.18456e-07,
+      "loss": 0.5251,
+      "step": 20400
+    },
+    {
+      "epoch": 0.2045,
+      "grad_norm": 2.283958911895752,
+      "learning_rate": 9.182559999999999e-07,
+      "loss": 0.4597,
+      "step": 20450
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 51.34020233154297,
+      "learning_rate": 9.180559999999999e-07,
+      "loss": 0.6181,
+      "step": 20500
+    },
+    {
+      "epoch": 0.2055,
+      "grad_norm": 113.1830062866211,
+      "learning_rate": 9.17856e-07,
+      "loss": 0.5824,
+      "step": 20550
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 95.66886901855469,
+      "learning_rate": 9.17656e-07,
+      "loss": 0.5657,
+      "step": 20600
+    },
+    {
+      "epoch": 0.2065,
+      "grad_norm": 72.63542938232422,
+      "learning_rate": 9.174560000000001e-07,
+      "loss": 0.5979,
+      "step": 20650
+    },
+    {
+      "epoch": 0.207,
+      "grad_norm": 62.32326889038086,
+      "learning_rate": 9.172559999999999e-07,
+      "loss": 0.5072,
+      "step": 20700
+    },
+    {
+      "epoch": 0.2075,
+      "grad_norm": 59.557228088378906,
+      "learning_rate": 9.170559999999999e-07,
+      "loss": 0.4438,
+      "step": 20750
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 64.50997161865234,
+      "learning_rate": 9.16856e-07,
+      "loss": 0.6501,
+      "step": 20800
+    },
+    {
+      "epoch": 0.2085,
+      "grad_norm": 51.724708557128906,
+      "learning_rate": 9.16656e-07,
+      "loss": 0.6389,
+      "step": 20850
+    },
+    {
+      "epoch": 0.209,
+      "grad_norm": 44.57424545288086,
+      "learning_rate": 9.16456e-07,
+      "loss": 0.5444,
+      "step": 20900
+    },
+    {
+      "epoch": 0.2095,
+      "grad_norm": 46.1285400390625,
+      "learning_rate": 9.16256e-07,
+      "loss": 0.6703,
+      "step": 20950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 62.80519485473633,
+      "learning_rate": 9.16056e-07,
+      "loss": 0.5194,
+      "step": 21000
+    },
+    {
+      "epoch": 0.2105,
+      "grad_norm": 97.80638885498047,
+      "learning_rate": 9.158559999999999e-07,
+      "loss": 0.425,
+      "step": 21050
+    },
+    {
+      "epoch": 0.211,
+      "grad_norm": 79.29085540771484,
+      "learning_rate": 9.15656e-07,
+      "loss": 0.4748,
+      "step": 21100
+    },
+    {
+      "epoch": 0.2115,
+      "grad_norm": 15.490723609924316,
+      "learning_rate": 9.15456e-07,
+      "loss": 0.666,
+      "step": 21150
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 1.4117364883422852,
+      "learning_rate": 9.152559999999999e-07,
+      "loss": 0.4462,
+      "step": 21200
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 64.279541015625,
+      "learning_rate": 9.15056e-07,
+      "loss": 0.462,
+      "step": 21250
+    },
+    {
+      "epoch": 0.213,
+      "grad_norm": 50.567176818847656,
+      "learning_rate": 9.14856e-07,
+      "loss": 0.629,
+      "step": 21300
+    },
+    {
+      "epoch": 0.2135,
+      "grad_norm": 74.72473907470703,
+      "learning_rate": 9.14656e-07,
+      "loss": 0.7797,
+      "step": 21350
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 0.2427138090133667,
+      "learning_rate": 9.14456e-07,
+      "loss": 0.6424,
+      "step": 21400
+    },
+    {
+      "epoch": 0.2145,
+      "grad_norm": 94.19923400878906,
+      "learning_rate": 9.142559999999999e-07,
+      "loss": 0.5077,
+      "step": 21450
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 5.564889430999756,
+      "learning_rate": 9.140559999999999e-07,
+      "loss": 0.5594,
+      "step": 21500
+    },
+    {
+      "epoch": 0.2155,
+      "grad_norm": 79.15170288085938,
+      "learning_rate": 9.1386e-07,
+      "loss": 0.5204,
+      "step": 21550
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 34.72391128540039,
+      "learning_rate": 9.1366e-07,
+      "loss": 0.5304,
+      "step": 21600
+    },
+    {
+      "epoch": 0.2165,
+      "grad_norm": 45.1458625793457,
+      "learning_rate": 9.1346e-07,
+      "loss": 0.5573,
+      "step": 21650
+    },
+    {
+      "epoch": 0.217,
+      "grad_norm": 13.129668235778809,
+      "learning_rate": 9.1326e-07,
+      "loss": 0.5866,
+      "step": 21700
+    },
+    {
+      "epoch": 0.2175,
+      "grad_norm": 94.04682159423828,
+      "learning_rate": 9.130599999999999e-07,
+      "loss": 0.5363,
+      "step": 21750
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 3.027968645095825,
+      "learning_rate": 9.128599999999999e-07,
+      "loss": 0.6121,
+      "step": 21800
+    },
+    {
+      "epoch": 0.2185,
+      "grad_norm": 14.584113121032715,
+      "learning_rate": 9.1266e-07,
+      "loss": 0.6755,
+      "step": 21850
+    },
+    {
+      "epoch": 0.219,
+      "grad_norm": 90.84115600585938,
+      "learning_rate": 9.1246e-07,
+      "loss": 0.8153,
+      "step": 21900
+    },
+    {
+      "epoch": 0.2195,
+      "grad_norm": 54.05946350097656,
+      "learning_rate": 9.122599999999999e-07,
+      "loss": 0.5227,
+      "step": 21950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 112.29281616210938,
+      "learning_rate": 9.1206e-07,
+      "loss": 0.5026,
+      "step": 22000
+    },
+    {
+      "epoch": 0.2205,
+      "grad_norm": 39.67966842651367,
+      "learning_rate": 9.1186e-07,
+      "loss": 0.6301,
+      "step": 22050
+    },
+    {
+      "epoch": 0.221,
+      "grad_norm": 80.91326141357422,
+      "learning_rate": 9.116599999999999e-07,
+      "loss": 0.5802,
+      "step": 22100
+    },
+    {
+      "epoch": 0.2215,
+      "grad_norm": 21.68891143798828,
+      "learning_rate": 9.1146e-07,
+      "loss": 0.6474,
+      "step": 22150
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 112.01923370361328,
+      "learning_rate": 9.112599999999999e-07,
+      "loss": 0.556,
+      "step": 22200
+    },
+    {
+      "epoch": 0.2225,
+      "grad_norm": 7.433606147766113,
+      "learning_rate": 9.110599999999999e-07,
+      "loss": 0.6761,
+      "step": 22250
+    },
+    {
+      "epoch": 0.223,
+      "grad_norm": 88.78018188476562,
+      "learning_rate": 9.1086e-07,
+      "loss": 0.6357,
+      "step": 22300
+    },
+    {
+      "epoch": 0.2235,
+      "grad_norm": 143.90103149414062,
+      "learning_rate": 9.1066e-07,
+      "loss": 0.5232,
+      "step": 22350
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 7.7126359939575195,
+      "learning_rate": 9.1046e-07,
+      "loss": 0.6542,
+      "step": 22400
+    },
+    {
+      "epoch": 0.2245,
+      "grad_norm": 44.35465621948242,
+      "learning_rate": 9.102599999999999e-07,
+      "loss": 0.7115,
+      "step": 22450
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 7.780481815338135,
+      "learning_rate": 9.100599999999999e-07,
+      "loss": 0.6885,
+      "step": 22500
+    },
+    {
+      "epoch": 0.2255,
+      "grad_norm": 120.73460388183594,
+      "learning_rate": 9.098599999999999e-07,
+      "loss": 0.6542,
+      "step": 22550
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 85.5213623046875,
+      "learning_rate": 9.0966e-07,
+      "loss": 0.6606,
+      "step": 22600
+    },
+    {
+      "epoch": 0.2265,
+      "grad_norm": 73.37017822265625,
+      "learning_rate": 9.0946e-07,
+      "loss": 0.6267,
+      "step": 22650
+    },
+    {
+      "epoch": 0.227,
+      "grad_norm": 89.31539916992188,
+      "learning_rate": 9.092599999999999e-07,
+      "loss": 0.7014,
+      "step": 22700
+    },
+    {
+      "epoch": 0.2275,
+      "grad_norm": 27.36491584777832,
+      "learning_rate": 9.0906e-07,
+      "loss": 0.6383,
+      "step": 22750
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 57.39030838012695,
+      "learning_rate": 9.0886e-07,
+      "loss": 0.6104,
+      "step": 22800
+    },
+    {
+      "epoch": 0.2285,
+      "grad_norm": 6.634609222412109,
+      "learning_rate": 9.086599999999999e-07,
+      "loss": 0.4772,
+      "step": 22850
+    },
+    {
+      "epoch": 0.229,
+      "grad_norm": 98.5889892578125,
+      "learning_rate": 9.0846e-07,
+      "loss": 0.6115,
+      "step": 22900
+    },
+    {
+      "epoch": 0.2295,
+      "grad_norm": 71.15877532958984,
+      "learning_rate": 9.082599999999999e-07,
+      "loss": 0.4788,
+      "step": 22950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 61.014102935791016,
+      "learning_rate": 9.080599999999999e-07,
+      "loss": 0.5738,
+      "step": 23000
+    },
+    {
+      "epoch": 0.2305,
+      "grad_norm": 18.213537216186523,
+      "learning_rate": 9.0786e-07,
+      "loss": 0.5771,
+      "step": 23050
+    },
+    {
+      "epoch": 0.231,
+      "grad_norm": 62.87067794799805,
+      "learning_rate": 9.0766e-07,
+      "loss": 0.5029,
+      "step": 23100
+    },
+    {
+      "epoch": 0.2315,
+      "grad_norm": 52.611976623535156,
+      "learning_rate": 9.074600000000001e-07,
+      "loss": 0.5436,
+      "step": 23150
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 79.97590637207031,
+      "learning_rate": 9.072599999999999e-07,
+      "loss": 0.5754,
+      "step": 23200
+    },
+    {
+      "epoch": 0.2325,
+      "grad_norm": 100.50433349609375,
+      "learning_rate": 9.070599999999999e-07,
+      "loss": 0.7087,
+      "step": 23250
+    },
+    {
+      "epoch": 0.233,
+      "grad_norm": 77.63613891601562,
+      "learning_rate": 9.0686e-07,
+      "loss": 0.4804,
+      "step": 23300
+    },
+    {
+      "epoch": 0.2335,
+      "grad_norm": 152.4717254638672,
+      "learning_rate": 9.0666e-07,
+      "loss": 0.4687,
+      "step": 23350
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 1.093911051750183,
+      "learning_rate": 9.0646e-07,
+      "loss": 0.5668,
+      "step": 23400
+    },
+    {
+      "epoch": 0.2345,
+      "grad_norm": 80.12809753417969,
+      "learning_rate": 9.0626e-07,
+      "loss": 0.6275,
+      "step": 23450
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 17.399316787719727,
+      "learning_rate": 9.0606e-07,
+      "loss": 0.5775,
+      "step": 23500
+    },
+    {
+      "epoch": 0.2355,
+      "grad_norm": 62.769813537597656,
+      "learning_rate": 9.058599999999999e-07,
+      "loss": 0.5202,
+      "step": 23550
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 37.83443069458008,
+      "learning_rate": 9.0566e-07,
+      "loss": 0.4658,
+      "step": 23600
+    },
+    {
+      "epoch": 0.2365,
+      "grad_norm": 16.844783782958984,
+      "learning_rate": 9.0546e-07,
+      "loss": 0.4983,
+      "step": 23650
+    },
+    {
+      "epoch": 0.237,
+      "grad_norm": 56.658695220947266,
+      "learning_rate": 9.052599999999999e-07,
+      "loss": 0.715,
+      "step": 23700
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 52.001991271972656,
+      "learning_rate": 9.0506e-07,
+      "loss": 0.481,
+      "step": 23750
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 44.0219612121582,
+      "learning_rate": 9.0486e-07,
+      "loss": 0.5652,
+      "step": 23800
+    },
+    {
+      "epoch": 0.2385,
+      "grad_norm": 23.275821685791016,
+      "learning_rate": 9.0466e-07,
+      "loss": 0.5805,
+      "step": 23850
+    },
+    {
+      "epoch": 0.239,
+      "grad_norm": 42.84242630004883,
+      "learning_rate": 9.0446e-07,
+      "loss": 0.579,
+      "step": 23900
+    },
+    {
+      "epoch": 0.2395,
+      "grad_norm": 38.506683349609375,
+      "learning_rate": 9.042599999999999e-07,
+      "loss": 0.6933,
+      "step": 23950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 97.8546142578125,
+      "learning_rate": 9.040599999999999e-07,
+      "loss": 0.5293,
+      "step": 24000
+    },
+    {
+      "epoch": 0.2405,
+      "grad_norm": 34.26162338256836,
+      "learning_rate": 9.0386e-07,
+      "loss": 0.6291,
+      "step": 24050
+    },
+    {
+      "epoch": 0.241,
+      "grad_norm": 52.09344482421875,
+      "learning_rate": 9.0366e-07,
+      "loss": 0.5158,
+      "step": 24100
+    },
+    {
+      "epoch": 0.2415,
+      "grad_norm": 52.938385009765625,
+      "learning_rate": 9.0346e-07,
+      "loss": 0.4271,
+      "step": 24150
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 102.01466369628906,
+      "learning_rate": 9.03264e-07,
+      "loss": 0.676,
+      "step": 24200
+    },
+    {
+      "epoch": 0.2425,
+      "grad_norm": 92.10142517089844,
+      "learning_rate": 9.030639999999999e-07,
+      "loss": 0.4368,
+      "step": 24250
+    },
+    {
+      "epoch": 0.243,
+      "grad_norm": 61.187618255615234,
+      "learning_rate": 9.028639999999999e-07,
+      "loss": 0.593,
+      "step": 24300
+    },
+    {
+      "epoch": 0.2435,
+      "grad_norm": 54.55991744995117,
+      "learning_rate": 9.02664e-07,
+      "loss": 0.5802,
+      "step": 24350
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 66.34545135498047,
+      "learning_rate": 9.02464e-07,
+      "loss": 0.5796,
+      "step": 24400
+    },
+    {
+      "epoch": 0.2445,
+      "grad_norm": 57.91377639770508,
+      "learning_rate": 9.022639999999999e-07,
+      "loss": 0.5566,
+      "step": 24450
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 40.273277282714844,
+      "learning_rate": 9.02064e-07,
+      "loss": 0.5867,
+      "step": 24500
+    },
+    {
+      "epoch": 0.2455,
+      "grad_norm": 83.97872161865234,
+      "learning_rate": 9.01864e-07,
+      "loss": 0.6223,
+      "step": 24550
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 50.599708557128906,
+      "learning_rate": 9.016639999999999e-07,
+      "loss": 0.498,
+      "step": 24600
+    },
+    {
+      "epoch": 0.2465,
+      "grad_norm": 14.626625061035156,
+      "learning_rate": 9.01464e-07,
+      "loss": 0.5453,
+      "step": 24650
+    },
+    {
+      "epoch": 0.247,
+      "grad_norm": 12.880644798278809,
+      "learning_rate": 9.012639999999999e-07,
+      "loss": 0.5644,
+      "step": 24700
+    },
+    {
+      "epoch": 0.2475,
+      "grad_norm": 10.257484436035156,
+      "learning_rate": 9.010639999999999e-07,
+      "loss": 0.4375,
+      "step": 24750
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 96.48957061767578,
+      "learning_rate": 9.00864e-07,
+      "loss": 0.6164,
+      "step": 24800
+    },
+    {
+      "epoch": 0.2485,
+      "grad_norm": 75.5651626586914,
+      "learning_rate": 9.00664e-07,
+      "loss": 0.82,
+      "step": 24850
+    },
+    {
+      "epoch": 0.249,
+      "grad_norm": 29.0128116607666,
+      "learning_rate": 9.00464e-07,
+      "loss": 0.4682,
+      "step": 24900
+    },
+    {
+      "epoch": 0.2495,
+      "grad_norm": 15.282790184020996,
+      "learning_rate": 9.002639999999999e-07,
+      "loss": 0.512,
+      "step": 24950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 70.66602325439453,
+      "learning_rate": 9.000639999999999e-07,
+      "loss": 0.719,
+      "step": 25000
+    },
+    {
+      "epoch": 0.2505,
+      "grad_norm": 133.6775665283203,
+      "learning_rate": 8.998639999999999e-07,
+      "loss": 0.583,
+      "step": 25050
+    },
+    {
+      "epoch": 0.251,
+      "grad_norm": 53.346988677978516,
+      "learning_rate": 8.99664e-07,
+      "loss": 0.661,
+      "step": 25100
+    },
+    {
+      "epoch": 0.2515,
+      "grad_norm": 57.71779251098633,
+      "learning_rate": 8.99464e-07,
+      "loss": 0.4995,
+      "step": 25150
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 11.651928901672363,
+      "learning_rate": 8.992639999999999e-07,
+      "loss": 0.5597,
+      "step": 25200
+    },
+    {
+      "epoch": 0.2525,
+      "grad_norm": 107.9719467163086,
+      "learning_rate": 8.99064e-07,
+      "loss": 0.497,
+      "step": 25250
+    },
+    {
+      "epoch": 0.253,
+      "grad_norm": 14.6730375289917,
+      "learning_rate": 8.988639999999999e-07,
+      "loss": 0.6174,
+      "step": 25300
+    },
+    {
+      "epoch": 0.2535,
+      "grad_norm": 56.200225830078125,
+      "learning_rate": 8.986639999999999e-07,
+      "loss": 0.6374,
+      "step": 25350
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 15.914813995361328,
+      "learning_rate": 8.98464e-07,
+      "loss": 0.513,
+      "step": 25400
+    },
+    {
+      "epoch": 0.2545,
+      "grad_norm": 3.805548906326294,
+      "learning_rate": 8.982639999999999e-07,
+      "loss": 0.5197,
+      "step": 25450
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 88.02377319335938,
+      "learning_rate": 8.980639999999999e-07,
+      "loss": 0.7373,
+      "step": 25500
+    },
+    {
+      "epoch": 0.2555,
+      "grad_norm": 35.10052490234375,
+      "learning_rate": 8.97864e-07,
+      "loss": 0.626,
+      "step": 25550
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 40.560298919677734,
+      "learning_rate": 8.97664e-07,
+      "loss": 0.6376,
+      "step": 25600
+    },
+    {
+      "epoch": 0.2565,
+      "grad_norm": 36.76740264892578,
+      "learning_rate": 8.97464e-07,
+      "loss": 0.6143,
+      "step": 25650
+    },
+    {
+      "epoch": 0.257,
+      "grad_norm": 58.6038703918457,
+      "learning_rate": 8.972639999999999e-07,
+      "loss": 0.6131,
+      "step": 25700
+    },
+    {
+      "epoch": 0.2575,
+      "grad_norm": 16.4853515625,
+      "learning_rate": 8.970639999999999e-07,
+      "loss": 0.5266,
+      "step": 25750
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 2.5067620277404785,
+      "learning_rate": 8.96864e-07,
+      "loss": 0.4183,
+      "step": 25800
+    },
+    {
+      "epoch": 0.2585,
+      "grad_norm": 47.051795959472656,
+      "learning_rate": 8.96664e-07,
+      "loss": 0.411,
+      "step": 25850
+    },
+    {
+      "epoch": 0.259,
+      "grad_norm": 80.19731140136719,
+      "learning_rate": 8.96464e-07,
+      "loss": 0.6141,
+      "step": 25900
+    },
+    {
+      "epoch": 0.2595,
+      "grad_norm": 53.69868469238281,
+      "learning_rate": 8.96264e-07,
+      "loss": 0.5144,
+      "step": 25950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 33.09700393676758,
+      "learning_rate": 8.960639999999999e-07,
+      "loss": 0.6407,
+      "step": 26000
+    },
+    {
+      "epoch": 0.2605,
+      "grad_norm": 28.68010711669922,
+      "learning_rate": 8.958639999999999e-07,
+      "loss": 0.4799,
+      "step": 26050
+    },
+    {
+      "epoch": 0.261,
+      "grad_norm": 31.614540100097656,
+      "learning_rate": 8.95664e-07,
+      "loss": 0.5902,
+      "step": 26100
+    },
+    {
+      "epoch": 0.2615,
+      "grad_norm": 46.6901969909668,
+      "learning_rate": 8.95464e-07,
+      "loss": 0.5856,
+      "step": 26150
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 22.507549285888672,
+      "learning_rate": 8.952639999999999e-07,
+      "loss": 0.6191,
+      "step": 26200
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 25.52752113342285,
+      "learning_rate": 8.95064e-07,
+      "loss": 0.5374,
+      "step": 26250
+    },
+    {
+      "epoch": 0.263,
+      "grad_norm": 66.70597076416016,
+      "learning_rate": 8.94864e-07,
+      "loss": 0.5224,
+      "step": 26300
+    },
+    {
+      "epoch": 0.2635,
+      "grad_norm": 62.320648193359375,
+      "learning_rate": 8.946639999999999e-07,
+      "loss": 0.5897,
+      "step": 26350
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 15.186070442199707,
+      "learning_rate": 8.94464e-07,
+      "loss": 0.6228,
+      "step": 26400
+    },
+    {
+      "epoch": 0.2645,
+      "grad_norm": 111.6026382446289,
+      "learning_rate": 8.942639999999999e-07,
+      "loss": 0.5744,
+      "step": 26450
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 82.44178771972656,
+      "learning_rate": 8.940639999999999e-07,
+      "loss": 0.5254,
+      "step": 26500
+    },
+    {
+      "epoch": 0.2655,
+      "grad_norm": 52.06359100341797,
+      "learning_rate": 8.93864e-07,
+      "loss": 0.4563,
+      "step": 26550
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 77.35059356689453,
+      "learning_rate": 8.93664e-07,
+      "loss": 0.5443,
+      "step": 26600
+    },
+    {
+      "epoch": 0.2665,
+      "grad_norm": 40.24097442626953,
+      "learning_rate": 8.93464e-07,
+      "loss": 0.7312,
+      "step": 26650
+    },
+    {
+      "epoch": 0.267,
+      "grad_norm": 12.489123344421387,
+      "learning_rate": 8.932639999999999e-07,
+      "loss": 0.5649,
+      "step": 26700
+    },
+    {
+      "epoch": 0.2675,
+      "grad_norm": 87.02991485595703,
+      "learning_rate": 8.930639999999999e-07,
+      "loss": 0.5359,
+      "step": 26750
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 20.647361755371094,
+      "learning_rate": 8.928639999999999e-07,
+      "loss": 0.5762,
+      "step": 26800
+    },
+    {
+      "epoch": 0.2685,
+      "grad_norm": 67.69243621826172,
+      "learning_rate": 8.92664e-07,
+      "loss": 0.4922,
+      "step": 26850
+    },
+    {
+      "epoch": 0.269,
+      "grad_norm": 60.18714904785156,
+      "learning_rate": 8.92464e-07,
+      "loss": 0.5516,
+      "step": 26900
+    },
+    {
+      "epoch": 0.2695,
+      "grad_norm": 9.904438972473145,
+      "learning_rate": 8.922639999999999e-07,
+      "loss": 0.5436,
+      "step": 26950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 74.50110626220703,
+      "learning_rate": 8.92064e-07,
+      "loss": 0.6846,
+      "step": 27000
+    },
+    {
+      "epoch": 0.2705,
+      "grad_norm": 64.5053939819336,
+      "learning_rate": 8.918639999999999e-07,
+      "loss": 0.5902,
+      "step": 27050
+    },
+    {
+      "epoch": 0.271,
+      "grad_norm": 54.11682891845703,
+      "learning_rate": 8.916679999999999e-07,
+      "loss": 0.6153,
+      "step": 27100
+    },
+    {
+      "epoch": 0.2715,
+      "grad_norm": 2.7656333446502686,
+      "learning_rate": 8.91468e-07,
+      "loss": 0.4774,
+      "step": 27150
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 47.780067443847656,
+      "learning_rate": 8.912679999999999e-07,
+      "loss": 0.6265,
+      "step": 27200
+    },
+    {
+      "epoch": 0.2725,
+      "grad_norm": 79.4407958984375,
+      "learning_rate": 8.910679999999999e-07,
+      "loss": 0.4763,
+      "step": 27250
+    },
+    {
+      "epoch": 0.273,
+      "grad_norm": 30.775165557861328,
+      "learning_rate": 8.90868e-07,
+      "loss": 0.4733,
+      "step": 27300
+    },
+    {
+      "epoch": 0.2735,
+      "grad_norm": 43.01605987548828,
+      "learning_rate": 8.90668e-07,
+      "loss": 0.5996,
+      "step": 27350
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 25.7813663482666,
+      "learning_rate": 8.90468e-07,
+      "loss": 0.4017,
+      "step": 27400
+    },
+    {
+      "epoch": 0.2745,
+      "grad_norm": 70.08576202392578,
+      "learning_rate": 8.902679999999999e-07,
+      "loss": 0.6638,
+      "step": 27450
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 3.8417301177978516,
+      "learning_rate": 8.900679999999999e-07,
+      "loss": 0.5934,
+      "step": 27500
+    },
+    {
+      "epoch": 0.2755,
+      "grad_norm": 43.573211669921875,
+      "learning_rate": 8.898679999999999e-07,
+      "loss": 0.4682,
+      "step": 27550
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 77.05116271972656,
+      "learning_rate": 8.89668e-07,
+      "loss": 0.852,
+      "step": 27600
+    },
+    {
+      "epoch": 0.2765,
+      "grad_norm": 51.25703430175781,
+      "learning_rate": 8.89468e-07,
+      "loss": 0.5171,
+      "step": 27650
+    },
+    {
+      "epoch": 0.277,
+      "grad_norm": 79.16458129882812,
+      "learning_rate": 8.892679999999999e-07,
+      "loss": 0.4679,
+      "step": 27700
+    },
+    {
+      "epoch": 0.2775,
+      "grad_norm": 52.861976623535156,
+      "learning_rate": 8.89068e-07,
+      "loss": 0.5379,
+      "step": 27750
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 56.5855598449707,
+      "learning_rate": 8.888679999999999e-07,
+      "loss": 0.5315,
+      "step": 27800
+    },
+    {
+      "epoch": 0.2785,
+      "grad_norm": 38.6595458984375,
+      "learning_rate": 8.886679999999999e-07,
+      "loss": 0.5684,
+      "step": 27850
+    },
+    {
+      "epoch": 0.279,
+      "grad_norm": 111.93008422851562,
+      "learning_rate": 8.88468e-07,
+      "loss": 0.5945,
+      "step": 27900
+    },
+    {
+      "epoch": 0.2795,
+      "grad_norm": 58.56281661987305,
+      "learning_rate": 8.882679999999999e-07,
+      "loss": 0.4908,
+      "step": 27950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.173264026641846,
+      "learning_rate": 8.880679999999999e-07,
+      "loss": 0.6168,
+      "step": 28000
+    },
+    {
+      "epoch": 0.2805,
+      "grad_norm": 75.64866638183594,
+      "learning_rate": 8.87868e-07,
+      "loss": 0.6355,
+      "step": 28050
+    },
+    {
+      "epoch": 0.281,
+      "grad_norm": 17.418989181518555,
+      "learning_rate": 8.87668e-07,
+      "loss": 0.4972,
+      "step": 28100
+    },
+    {
+      "epoch": 0.2815,
+      "grad_norm": 105.94364929199219,
+      "learning_rate": 8.87468e-07,
+      "loss": 0.4307,
+      "step": 28150
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 204.995361328125,
+      "learning_rate": 8.872679999999999e-07,
+      "loss": 0.5594,
+      "step": 28200
+    },
+    {
+      "epoch": 0.2825,
+      "grad_norm": 64.4760971069336,
+      "learning_rate": 8.870679999999999e-07,
+      "loss": 0.7741,
+      "step": 28250
+    },
+    {
+      "epoch": 0.283,
+      "grad_norm": 88.27828216552734,
+      "learning_rate": 8.86868e-07,
+      "loss": 0.6538,
+      "step": 28300
+    },
+    {
+      "epoch": 0.2835,
+      "grad_norm": 16.012582778930664,
+      "learning_rate": 8.86668e-07,
+      "loss": 0.6104,
+      "step": 28350
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 27.557018280029297,
+      "learning_rate": 8.86468e-07,
+      "loss": 0.501,
+      "step": 28400
+    },
+    {
+      "epoch": 0.2845,
+      "grad_norm": 54.364566802978516,
+      "learning_rate": 8.86268e-07,
+      "loss": 0.5534,
+      "step": 28450
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 17.52170753479004,
+      "learning_rate": 8.860679999999999e-07,
+      "loss": 0.5886,
+      "step": 28500
+    },
+    {
+      "epoch": 0.2855,
+      "grad_norm": 65.14887237548828,
+      "learning_rate": 8.858679999999999e-07,
+      "loss": 0.5129,
+      "step": 28550
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 4.750748157501221,
+      "learning_rate": 8.85668e-07,
+      "loss": 0.521,
+      "step": 28600
+    },
+    {
+      "epoch": 0.2865,
+      "grad_norm": 69.85326385498047,
+      "learning_rate": 8.85468e-07,
+      "loss": 0.5333,
+      "step": 28650
+    },
+    {
+      "epoch": 0.287,
+      "grad_norm": 54.12617874145508,
+      "learning_rate": 8.85268e-07,
+      "loss": 0.481,
+      "step": 28700
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 1.6394026279449463,
+      "learning_rate": 8.85068e-07,
+      "loss": 0.5055,
+      "step": 28750
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 12.818920135498047,
+      "learning_rate": 8.84868e-07,
+      "loss": 0.6489,
+      "step": 28800
+    },
+    {
+      "epoch": 0.2885,
+      "grad_norm": 29.769569396972656,
+      "learning_rate": 8.846679999999999e-07,
+      "loss": 0.4749,
+      "step": 28850
+    },
+    {
+      "epoch": 0.289,
+      "grad_norm": 62.24850845336914,
+      "learning_rate": 8.84468e-07,
+      "loss": 0.5205,
+      "step": 28900
+    },
+    {
+      "epoch": 0.2895,
+      "grad_norm": 6.054821491241455,
+      "learning_rate": 8.84268e-07,
+      "loss": 0.5548,
+      "step": 28950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 57.04275894165039,
+      "learning_rate": 8.840679999999999e-07,
+      "loss": 0.435,
+      "step": 29000
+    },
+    {
+      "epoch": 0.2905,
+      "grad_norm": 44.01559829711914,
+      "learning_rate": 8.83868e-07,
+      "loss": 0.5515,
+      "step": 29050
+    },
+    {
+      "epoch": 0.291,
+      "grad_norm": 102.21464538574219,
+      "learning_rate": 8.83668e-07,
+      "loss": 0.5175,
+      "step": 29100
+    },
+    {
+      "epoch": 0.2915,
+      "grad_norm": 16.0670108795166,
+      "learning_rate": 8.83468e-07,
+      "loss": 0.5667,
+      "step": 29150
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 50.337894439697266,
+      "learning_rate": 8.83268e-07,
+      "loss": 0.6014,
+      "step": 29200
+    },
+    {
+      "epoch": 0.2925,
+      "grad_norm": 90.58765411376953,
+      "learning_rate": 8.830679999999999e-07,
+      "loss": 0.5361,
+      "step": 29250
+    },
+    {
+      "epoch": 0.293,
+      "grad_norm": 88.60356140136719,
+      "learning_rate": 8.828679999999999e-07,
+      "loss": 0.4992,
+      "step": 29300
+    },
+    {
+      "epoch": 0.2935,
+      "grad_norm": 17.39057731628418,
+      "learning_rate": 8.82668e-07,
+      "loss": 0.6699,
+      "step": 29350
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 9.938417434692383,
+      "learning_rate": 8.82468e-07,
+      "loss": 0.5748,
+      "step": 29400
+    },
+    {
+      "epoch": 0.2945,
+      "grad_norm": 86.84083557128906,
+      "learning_rate": 8.82268e-07,
+      "loss": 0.5455,
+      "step": 29450
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 76.97161102294922,
+      "learning_rate": 8.82068e-07,
+      "loss": 0.5668,
+      "step": 29500
+    },
+    {
+      "epoch": 0.2955,
+      "grad_norm": 6.305858135223389,
+      "learning_rate": 8.818679999999999e-07,
+      "loss": 0.4598,
+      "step": 29550
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 83.46090698242188,
+      "learning_rate": 8.816679999999999e-07,
+      "loss": 0.5434,
+      "step": 29600
+    },
+    {
+      "epoch": 0.2965,
+      "grad_norm": 39.0470085144043,
+      "learning_rate": 8.81468e-07,
+      "loss": 0.5519,
+      "step": 29650
+    },
+    {
+      "epoch": 0.297,
+      "grad_norm": 74.89450073242188,
+      "learning_rate": 8.81268e-07,
+      "loss": 0.6435,
+      "step": 29700
+    },
+    {
+      "epoch": 0.2975,
+      "grad_norm": 77.68589782714844,
+      "learning_rate": 8.810679999999999e-07,
+      "loss": 0.5398,
+      "step": 29750
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 55.85251235961914,
+      "learning_rate": 8.80868e-07,
+      "loss": 0.5951,
+      "step": 29800
+    },
+    {
+      "epoch": 0.2985,
+      "grad_norm": 35.89079666137695,
+      "learning_rate": 8.80668e-07,
+      "loss": 0.5007,
+      "step": 29850
+    },
+    {
+      "epoch": 0.299,
+      "grad_norm": 81.23458099365234,
+      "learning_rate": 8.804679999999999e-07,
+      "loss": 0.6661,
+      "step": 29900
+    },
+    {
+      "epoch": 0.2995,
+      "grad_norm": 30.005859375,
+      "learning_rate": 8.80268e-07,
+      "loss": 0.5404,
+      "step": 29950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 93.9574203491211,
+      "learning_rate": 8.800679999999999e-07,
+      "loss": 0.5561,
+      "step": 30000
+    },
+    {
+      "epoch": 0.3005,
+      "grad_norm": 70.64046478271484,
+      "learning_rate": 8.798719999999999e-07,
+      "loss": 0.4643,
+      "step": 30050
+    },
+    {
+      "epoch": 0.301,
+      "grad_norm": 23.76947021484375,
+      "learning_rate": 8.79672e-07,
+      "loss": 0.5806,
+      "step": 30100
+    },
+    {
+      "epoch": 0.3015,
+      "grad_norm": 54.23265838623047,
+      "learning_rate": 8.79472e-07,
+      "loss": 0.5467,
+      "step": 30150
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 62.528175354003906,
+      "learning_rate": 8.792720000000001e-07,
+      "loss": 0.6381,
+      "step": 30200
+    },
+    {
+      "epoch": 0.3025,
+      "grad_norm": 24.852832794189453,
+      "learning_rate": 8.790719999999999e-07,
+      "loss": 0.4532,
+      "step": 30250
+    },
+    {
+      "epoch": 0.303,
+      "grad_norm": 55.6757698059082,
+      "learning_rate": 8.788719999999999e-07,
+      "loss": 0.7508,
+      "step": 30300
+    },
+    {
+      "epoch": 0.3035,
+      "grad_norm": 108.9383316040039,
+      "learning_rate": 8.78672e-07,
+      "loss": 0.6378,
+      "step": 30350
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 44.41104507446289,
+      "learning_rate": 8.78472e-07,
+      "loss": 0.6764,
+      "step": 30400
+    },
+    {
+      "epoch": 0.3045,
+      "grad_norm": 34.045413970947266,
+      "learning_rate": 8.78272e-07,
+      "loss": 0.5944,
+      "step": 30450
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 29.27425765991211,
+      "learning_rate": 8.78072e-07,
+      "loss": 0.5212,
+      "step": 30500
+    },
+    {
+      "epoch": 0.3055,
+      "grad_norm": 83.5364761352539,
+      "learning_rate": 8.77872e-07,
+      "loss": 0.3991,
+      "step": 30550
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 1.8257538080215454,
+      "learning_rate": 8.776719999999999e-07,
+      "loss": 0.5711,
+      "step": 30600
+    },
+    {
+      "epoch": 0.3065,
+      "grad_norm": 74.53131866455078,
+      "learning_rate": 8.77472e-07,
+      "loss": 0.5351,
+      "step": 30650
+    },
+    {
+      "epoch": 0.307,
+      "grad_norm": 49.41853713989258,
+      "learning_rate": 8.77272e-07,
+      "loss": 0.5694,
+      "step": 30700
+    },
+    {
+      "epoch": 0.3075,
+      "grad_norm": 120.75214385986328,
+      "learning_rate": 8.770719999999999e-07,
+      "loss": 0.5498,
+      "step": 30750
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 46.72524642944336,
+      "learning_rate": 8.76872e-07,
+      "loss": 0.5695,
+      "step": 30800
+    },
+    {
+      "epoch": 0.3085,
+      "grad_norm": 88.22650146484375,
+      "learning_rate": 8.76672e-07,
+      "loss": 0.5629,
+      "step": 30850
+    },
+    {
+      "epoch": 0.309,
+      "grad_norm": 11.028036117553711,
+      "learning_rate": 8.76472e-07,
+      "loss": 0.4899,
+      "step": 30900
+    },
+    {
+      "epoch": 0.3095,
+      "grad_norm": 18.966903686523438,
+      "learning_rate": 8.76272e-07,
+      "loss": 0.5321,
+      "step": 30950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.517266273498535,
+      "learning_rate": 8.760719999999999e-07,
+      "loss": 0.5198,
+      "step": 31000
+    },
+    {
+      "epoch": 0.3105,
+      "grad_norm": 2.6922929286956787,
+      "learning_rate": 8.758719999999999e-07,
+      "loss": 0.4756,
+      "step": 31050
+    },
+    {
+      "epoch": 0.311,
+      "grad_norm": 80.45319366455078,
+      "learning_rate": 8.75672e-07,
+      "loss": 0.625,
+      "step": 31100
+    },
+    {
+      "epoch": 0.3115,
+      "grad_norm": 50.82831573486328,
+      "learning_rate": 8.75472e-07,
+      "loss": 0.5731,
+      "step": 31150
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 80.17171478271484,
+      "learning_rate": 8.75272e-07,
+      "loss": 0.4188,
+      "step": 31200
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 83.52489471435547,
+      "learning_rate": 8.75072e-07,
+      "loss": 0.6047,
+      "step": 31250
+    },
+    {
+      "epoch": 0.313,
+      "grad_norm": 40.45694351196289,
+      "learning_rate": 8.74872e-07,
+      "loss": 0.5314,
+      "step": 31300
+    },
+    {
+      "epoch": 0.3135,
+      "grad_norm": 7.2418694496154785,
+      "learning_rate": 8.746719999999999e-07,
+      "loss": 0.6789,
+      "step": 31350
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 105.38126373291016,
+      "learning_rate": 8.74472e-07,
+      "loss": 0.6344,
+      "step": 31400
+    },
+    {
+      "epoch": 0.3145,
+      "grad_norm": 69.45547485351562,
+      "learning_rate": 8.74272e-07,
+      "loss": 0.508,
+      "step": 31450
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 7.970687389373779,
+      "learning_rate": 8.740719999999999e-07,
+      "loss": 0.6869,
+      "step": 31500
+    },
+    {
+      "epoch": 0.3155,
+      "grad_norm": 85.00370788574219,
+      "learning_rate": 8.73872e-07,
+      "loss": 0.5232,
+      "step": 31550
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 63.42658996582031,
+      "learning_rate": 8.73672e-07,
+      "loss": 0.5965,
+      "step": 31600
+    },
+    {
+      "epoch": 0.3165,
+      "grad_norm": 100.0108871459961,
+      "learning_rate": 8.73472e-07,
+      "loss": 0.5686,
+      "step": 31650
+    },
+    {
+      "epoch": 0.317,
+      "grad_norm": 76.61862182617188,
+      "learning_rate": 8.73272e-07,
+      "loss": 0.6126,
+      "step": 31700
+    },
+    {
+      "epoch": 0.3175,
+      "grad_norm": 23.5545711517334,
+      "learning_rate": 8.730719999999999e-07,
+      "loss": 0.486,
+      "step": 31750
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 121.40791320800781,
+      "learning_rate": 8.728719999999999e-07,
+      "loss": 0.5634,
+      "step": 31800
+    },
+    {
+      "epoch": 0.3185,
+      "grad_norm": 60.57836151123047,
+      "learning_rate": 8.72672e-07,
+      "loss": 0.477,
+      "step": 31850
+    },
+    {
+      "epoch": 0.319,
+      "grad_norm": 71.34037017822266,
+      "learning_rate": 8.72472e-07,
+      "loss": 0.6271,
+      "step": 31900
+    },
+    {
+      "epoch": 0.3195,
+      "grad_norm": 1.094132661819458,
+      "learning_rate": 8.72272e-07,
+      "loss": 0.6245,
+      "step": 31950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 112.46796417236328,
+      "learning_rate": 8.72072e-07,
+      "loss": 0.4673,
+      "step": 32000
+    },
+    {
+      "epoch": 0.3205,
+      "grad_norm": 65.74776458740234,
+      "learning_rate": 8.718719999999999e-07,
+      "loss": 0.4702,
+      "step": 32050
+    },
+    {
+      "epoch": 0.321,
+      "grad_norm": 133.10072326660156,
+      "learning_rate": 8.716719999999999e-07,
+      "loss": 0.4708,
+      "step": 32100
+    },
+    {
+      "epoch": 0.3215,
+      "grad_norm": 88.02649688720703,
+      "learning_rate": 8.71472e-07,
+      "loss": 0.4592,
+      "step": 32150
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 97.91495513916016,
+      "learning_rate": 8.71272e-07,
+      "loss": 0.6828,
+      "step": 32200
+    },
+    {
+      "epoch": 0.3225,
+      "grad_norm": 76.7396011352539,
+      "learning_rate": 8.710719999999999e-07,
+      "loss": 0.4356,
+      "step": 32250
+    },
+    {
+      "epoch": 0.323,
+      "grad_norm": 48.60926055908203,
+      "learning_rate": 8.70872e-07,
+      "loss": 0.5898,
+      "step": 32300
+    },
+    {
+      "epoch": 0.3235,
+      "grad_norm": 107.64606475830078,
+      "learning_rate": 8.70672e-07,
+      "loss": 0.552,
+      "step": 32350
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 77.53285217285156,
+      "learning_rate": 8.704719999999999e-07,
+      "loss": 0.4998,
+      "step": 32400
+    },
+    {
+      "epoch": 0.3245,
+      "grad_norm": 3.892568349838257,
+      "learning_rate": 8.70272e-07,
+      "loss": 0.5265,
+      "step": 32450
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 87.63228607177734,
+      "learning_rate": 8.700719999999999e-07,
+      "loss": 0.6641,
+      "step": 32500
+    },
+    {
+      "epoch": 0.3255,
+      "grad_norm": 208.2560272216797,
+      "learning_rate": 8.698719999999999e-07,
+      "loss": 0.5142,
+      "step": 32550
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 39.32928466796875,
+      "learning_rate": 8.69672e-07,
+      "loss": 0.5655,
+      "step": 32600
+    },
+    {
+      "epoch": 0.3265,
+      "grad_norm": 121.70012664794922,
+      "learning_rate": 8.69472e-07,
+      "loss": 0.5668,
+      "step": 32650
+    },
+    {
+      "epoch": 0.327,
+      "grad_norm": 3.1223995685577393,
+      "learning_rate": 8.692720000000001e-07,
+      "loss": 0.5199,
+      "step": 32700
+    },
+    {
+      "epoch": 0.3275,
+      "grad_norm": 103.2892074584961,
+      "learning_rate": 8.690719999999999e-07,
+      "loss": 0.5335,
+      "step": 32750
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 9.383655548095703,
+      "learning_rate": 8.688719999999999e-07,
+      "loss": 0.6043,
+      "step": 32800
+    },
+    {
+      "epoch": 0.3285,
+      "grad_norm": 41.5776252746582,
+      "learning_rate": 8.68672e-07,
+      "loss": 0.671,
+      "step": 32850
+    },
+    {
+      "epoch": 0.329,
+      "grad_norm": 12.326675415039062,
+      "learning_rate": 8.68472e-07,
+      "loss": 0.4646,
+      "step": 32900
+    },
+    {
+      "epoch": 0.3295,
+      "grad_norm": 42.7502555847168,
+      "learning_rate": 8.68272e-07,
+      "loss": 0.492,
+      "step": 32950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 70.67731475830078,
+      "learning_rate": 8.68072e-07,
+      "loss": 0.4932,
+      "step": 33000
+    },
+    {
+      "epoch": 0.3305,
+      "grad_norm": 50.19448471069336,
+      "learning_rate": 8.67872e-07,
+      "loss": 0.4514,
+      "step": 33050
+    },
+    {
+      "epoch": 0.331,
+      "grad_norm": 73.63932037353516,
+      "learning_rate": 8.676719999999999e-07,
+      "loss": 0.4684,
+      "step": 33100
+    },
+    {
+      "epoch": 0.3315,
+      "grad_norm": 22.24913787841797,
+      "learning_rate": 8.67472e-07,
+      "loss": 0.6631,
+      "step": 33150
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 110.29122161865234,
+      "learning_rate": 8.67272e-07,
+      "loss": 0.714,
+      "step": 33200
+    },
+    {
+      "epoch": 0.3325,
+      "grad_norm": 9.803537368774414,
+      "learning_rate": 8.670719999999999e-07,
+      "loss": 0.5666,
+      "step": 33250
+    },
+    {
+      "epoch": 0.333,
+      "grad_norm": 1.9911659955978394,
+      "learning_rate": 8.66872e-07,
+      "loss": 0.5231,
+      "step": 33300
+    },
+    {
+      "epoch": 0.3335,
+      "grad_norm": 29.713672637939453,
+      "learning_rate": 8.66672e-07,
+      "loss": 0.6263,
+      "step": 33350
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 58.22745895385742,
+      "learning_rate": 8.66472e-07,
+      "loss": 0.4424,
+      "step": 33400
+    },
+    {
+      "epoch": 0.3345,
+      "grad_norm": 71.6273422241211,
+      "learning_rate": 8.66272e-07,
+      "loss": 0.4962,
+      "step": 33450
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 39.40334701538086,
+      "learning_rate": 8.660719999999999e-07,
+      "loss": 0.5212,
+      "step": 33500
+    },
+    {
+      "epoch": 0.3355,
+      "grad_norm": 100.38050842285156,
+      "learning_rate": 8.658719999999999e-07,
+      "loss": 0.432,
+      "step": 33550
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 57.89727783203125,
+      "learning_rate": 8.65672e-07,
+      "loss": 0.528,
+      "step": 33600
+    },
+    {
+      "epoch": 0.3365,
+      "grad_norm": 28.244213104248047,
+      "learning_rate": 8.65472e-07,
+      "loss": 0.5949,
+      "step": 33650
+    },
+    {
+      "epoch": 0.337,
+      "grad_norm": 57.93000030517578,
+      "learning_rate": 8.65272e-07,
+      "loss": 0.534,
+      "step": 33700
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 38.53422164916992,
+      "learning_rate": 8.65072e-07,
+      "loss": 0.6568,
+      "step": 33750
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 77.30050659179688,
+      "learning_rate": 8.648719999999999e-07,
+      "loss": 0.5894,
+      "step": 33800
+    },
+    {
+      "epoch": 0.3385,
+      "grad_norm": 99.1224594116211,
+      "learning_rate": 8.646719999999999e-07,
+      "loss": 0.4271,
+      "step": 33850
+    },
+    {
+      "epoch": 0.339,
+      "grad_norm": 10.921067237854004,
+      "learning_rate": 8.64472e-07,
+      "loss": 0.457,
+      "step": 33900
+    },
+    {
+      "epoch": 0.3395,
+      "grad_norm": 83.70796203613281,
+      "learning_rate": 8.64272e-07,
+      "loss": 0.5231,
+      "step": 33950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 81.03082275390625,
+      "learning_rate": 8.640719999999999e-07,
+      "loss": 0.6068,
+      "step": 34000
+    },
+    {
+      "epoch": 0.3405,
+      "grad_norm": 18.295440673828125,
+      "learning_rate": 8.63872e-07,
+      "loss": 0.5386,
+      "step": 34050
+    },
+    {
+      "epoch": 0.341,
+      "grad_norm": 79.17333221435547,
+      "learning_rate": 8.63676e-07,
+      "loss": 0.5819,
+      "step": 34100
+    },
+    {
+      "epoch": 0.3415,
+      "grad_norm": 59.78744888305664,
+      "learning_rate": 8.634759999999999e-07,
+      "loss": 0.6588,
+      "step": 34150
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 87.66487121582031,
+      "learning_rate": 8.63276e-07,
+      "loss": 0.4755,
+      "step": 34200
+    },
+    {
+      "epoch": 0.3425,
+      "grad_norm": 40.66081237792969,
+      "learning_rate": 8.630799999999999e-07,
+      "loss": 0.4778,
+      "step": 34250
+    },
+    {
+      "epoch": 0.343,
+      "grad_norm": 99.85968780517578,
+      "learning_rate": 8.628799999999999e-07,
+      "loss": 0.631,
+      "step": 34300
+    },
+    {
+      "epoch": 0.3435,
+      "grad_norm": 116.81835174560547,
+      "learning_rate": 8.6268e-07,
+      "loss": 0.58,
+      "step": 34350
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 20.3864803314209,
+      "learning_rate": 8.6248e-07,
+      "loss": 0.5837,
+      "step": 34400
+    },
+    {
+      "epoch": 0.3445,
+      "grad_norm": 22.575923919677734,
+      "learning_rate": 8.6228e-07,
+      "loss": 0.519,
+      "step": 34450
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 95.2472915649414,
+      "learning_rate": 8.6208e-07,
+      "loss": 0.5047,
+      "step": 34500
+    },
+    {
+      "epoch": 0.3455,
+      "grad_norm": 24.045143127441406,
+      "learning_rate": 8.618799999999999e-07,
+      "loss": 0.5282,
+      "step": 34550
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 93.23561096191406,
+      "learning_rate": 8.616799999999999e-07,
+      "loss": 0.5889,
+      "step": 34600
+    },
+    {
+      "epoch": 0.3465,
+      "grad_norm": 83.2427978515625,
+      "learning_rate": 8.6148e-07,
+      "loss": 0.395,
+      "step": 34650
+    },
+    {
+      "epoch": 0.347,
+      "grad_norm": 75.5633544921875,
+      "learning_rate": 8.6128e-07,
+      "loss": 0.5339,
+      "step": 34700
+    },
+    {
+      "epoch": 0.3475,
+      "grad_norm": 46.325660705566406,
+      "learning_rate": 8.610799999999999e-07,
+      "loss": 0.6007,
+      "step": 34750
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 94.44389343261719,
+      "learning_rate": 8.6088e-07,
+      "loss": 0.5271,
+      "step": 34800
+    },
+    {
+      "epoch": 0.3485,
+      "grad_norm": 1.9219623804092407,
+      "learning_rate": 8.6068e-07,
+      "loss": 0.4917,
+      "step": 34850
+    },
+    {
+      "epoch": 0.349,
+      "grad_norm": 256.146484375,
+      "learning_rate": 8.604799999999999e-07,
+      "loss": 0.4976,
+      "step": 34900
+    },
+    {
+      "epoch": 0.3495,
+      "grad_norm": 42.80830383300781,
+      "learning_rate": 8.6028e-07,
+      "loss": 0.4433,
+      "step": 34950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 38.0195426940918,
+      "learning_rate": 8.600799999999999e-07,
+      "loss": 0.615,
+      "step": 35000
+    },
+    {
+      "epoch": 0.3505,
+      "grad_norm": 55.40459060668945,
+      "learning_rate": 8.598799999999999e-07,
+      "loss": 0.5231,
+      "step": 35050
+    },
+    {
+      "epoch": 0.351,
+      "grad_norm": 75.99671173095703,
+      "learning_rate": 8.5968e-07,
+      "loss": 0.505,
+      "step": 35100
+    },
+    {
+      "epoch": 0.3515,
+      "grad_norm": 30.296842575073242,
+      "learning_rate": 8.5948e-07,
+      "loss": 0.5513,
+      "step": 35150
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 69.02685546875,
+      "learning_rate": 8.592800000000001e-07,
+      "loss": 0.585,
+      "step": 35200
+    },
+    {
+      "epoch": 0.3525,
+      "grad_norm": 88.14878845214844,
+      "learning_rate": 8.590799999999999e-07,
+      "loss": 0.6418,
+      "step": 35250
+    },
+    {
+      "epoch": 0.353,
+      "grad_norm": 1.586727261543274,
+      "learning_rate": 8.588799999999999e-07,
+      "loss": 0.5362,
+      "step": 35300
+    },
+    {
+      "epoch": 0.3535,
+      "grad_norm": 48.83518981933594,
+      "learning_rate": 8.5868e-07,
+      "loss": 0.4468,
+      "step": 35350
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 98.71307373046875,
+      "learning_rate": 8.5848e-07,
+      "loss": 0.6128,
+      "step": 35400
+    },
+    {
+      "epoch": 0.3545,
+      "grad_norm": 44.5135612487793,
+      "learning_rate": 8.5828e-07,
+      "loss": 0.695,
+      "step": 35450
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 39.28260803222656,
+      "learning_rate": 8.5808e-07,
+      "loss": 0.5119,
+      "step": 35500
+    },
+    {
+      "epoch": 0.3555,
+      "grad_norm": 16.02478790283203,
+      "learning_rate": 8.5788e-07,
+      "loss": 0.5504,
+      "step": 35550
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 73.70425415039062,
+      "learning_rate": 8.576799999999999e-07,
+      "loss": 0.3506,
+      "step": 35600
+    },
+    {
+      "epoch": 0.3565,
+      "grad_norm": 60.15646743774414,
+      "learning_rate": 8.5748e-07,
+      "loss": 0.701,
+      "step": 35650
+    },
+    {
+      "epoch": 0.357,
+      "grad_norm": 124.11194610595703,
+      "learning_rate": 8.5728e-07,
+      "loss": 0.5455,
+      "step": 35700
+    },
+    {
+      "epoch": 0.3575,
+      "grad_norm": 63.276222229003906,
+      "learning_rate": 8.570799999999999e-07,
+      "loss": 0.5035,
+      "step": 35750
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 70.08747863769531,
+      "learning_rate": 8.5688e-07,
+      "loss": 0.4866,
+      "step": 35800
+    },
+    {
+      "epoch": 0.3585,
+      "grad_norm": 85.71963500976562,
+      "learning_rate": 8.5668e-07,
+      "loss": 0.5919,
+      "step": 35850
+    },
+    {
+      "epoch": 0.359,
+      "grad_norm": 0.5785985589027405,
+      "learning_rate": 8.5648e-07,
+      "loss": 0.4748,
+      "step": 35900
+    },
+    {
+      "epoch": 0.3595,
+      "grad_norm": 54.46210861206055,
+      "learning_rate": 8.5628e-07,
+      "loss": 0.5546,
+      "step": 35950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6963469982147217,
+      "learning_rate": 8.560799999999999e-07,
+      "loss": 0.5244,
+      "step": 36000
+    },
+    {
+      "epoch": 0.3605,
+      "grad_norm": 75.58008575439453,
+      "learning_rate": 8.558799999999999e-07,
+      "loss": 0.493,
+      "step": 36050
+    },
+    {
+      "epoch": 0.361,
+      "grad_norm": 1.6295361518859863,
+      "learning_rate": 8.5568e-07,
+      "loss": 0.4227,
+      "step": 36100
+    },
+    {
+      "epoch": 0.3615,
+      "grad_norm": 78.0384292602539,
+      "learning_rate": 8.5548e-07,
+      "loss": 0.5416,
+      "step": 36150
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 14.516754150390625,
+      "learning_rate": 8.5528e-07,
+      "loss": 0.6593,
+      "step": 36200
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 29.74073028564453,
+      "learning_rate": 8.5508e-07,
+      "loss": 0.5811,
+      "step": 36250
+    },
+    {
+      "epoch": 0.363,
+      "grad_norm": 94.58187103271484,
+      "learning_rate": 8.548799999999999e-07,
+      "loss": 0.5598,
+      "step": 36300
+    },
+    {
+      "epoch": 0.3635,
+      "grad_norm": 71.66839599609375,
+      "learning_rate": 8.546799999999999e-07,
+      "loss": 0.3467,
+      "step": 36350
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 62.97394943237305,
+      "learning_rate": 8.5448e-07,
+      "loss": 0.668,
+      "step": 36400
+    },
+    {
+      "epoch": 0.3645,
+      "grad_norm": 24.426570892333984,
+      "learning_rate": 8.5428e-07,
+      "loss": 0.579,
+      "step": 36450
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 48.1230354309082,
+      "learning_rate": 8.540799999999999e-07,
+      "loss": 0.5345,
+      "step": 36500
+    },
+    {
+      "epoch": 0.3655,
+      "grad_norm": 82.27693176269531,
+      "learning_rate": 8.5388e-07,
+      "loss": 0.6415,
+      "step": 36550
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 29.432802200317383,
+      "learning_rate": 8.5368e-07,
+      "loss": 0.564,
+      "step": 36600
+    },
+    {
+      "epoch": 0.3665,
+      "grad_norm": 15.768404006958008,
+      "learning_rate": 8.534799999999999e-07,
+      "loss": 0.4265,
+      "step": 36650
+    },
+    {
+      "epoch": 0.367,
+      "grad_norm": 71.74403381347656,
+      "learning_rate": 8.5328e-07,
+      "loss": 0.5159,
+      "step": 36700
+    },
+    {
+      "epoch": 0.3675,
+      "grad_norm": 11.544425010681152,
+      "learning_rate": 8.530799999999999e-07,
+      "loss": 0.4023,
+      "step": 36750
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 6.664909839630127,
+      "learning_rate": 8.528799999999999e-07,
+      "loss": 0.6547,
+      "step": 36800
+    },
+    {
+      "epoch": 0.3685,
+      "grad_norm": 59.17173385620117,
+      "learning_rate": 8.5268e-07,
+      "loss": 0.4626,
+      "step": 36850
+    },
+    {
+      "epoch": 0.369,
+      "grad_norm": 45.66848373413086,
+      "learning_rate": 8.5248e-07,
+      "loss": 0.5634,
+      "step": 36900
+    },
+    {
+      "epoch": 0.3695,
+      "grad_norm": 34.434417724609375,
+      "learning_rate": 8.5228e-07,
+      "loss": 0.5269,
+      "step": 36950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 30.29810905456543,
+      "learning_rate": 8.520799999999999e-07,
+      "loss": 0.7069,
+      "step": 37000
+    },
+    {
+      "epoch": 0.3705,
+      "grad_norm": 75.58245849609375,
+      "learning_rate": 8.518799999999999e-07,
+      "loss": 0.5033,
+      "step": 37050
+    },
+    {
+      "epoch": 0.371,
+      "grad_norm": 1.9364076852798462,
+      "learning_rate": 8.516799999999999e-07,
+      "loss": 0.4915,
+      "step": 37100
+    },
+    {
+      "epoch": 0.3715,
+      "grad_norm": 48.733604431152344,
+      "learning_rate": 8.5148e-07,
+      "loss": 0.4742,
+      "step": 37150
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 98.77400970458984,
+      "learning_rate": 8.5128e-07,
+      "loss": 0.4848,
+      "step": 37200
+    },
+    {
+      "epoch": 0.3725,
+      "grad_norm": 126.41114044189453,
+      "learning_rate": 8.510799999999999e-07,
+      "loss": 0.4387,
+      "step": 37250
+    },
+    {
+      "epoch": 0.373,
+      "grad_norm": 93.77721405029297,
+      "learning_rate": 8.5088e-07,
+      "loss": 0.5441,
+      "step": 37300
+    },
+    {
+      "epoch": 0.3735,
+      "grad_norm": 22.67792510986328,
+      "learning_rate": 8.506799999999999e-07,
+      "loss": 0.4786,
+      "step": 37350
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 27.49308967590332,
+      "learning_rate": 8.504799999999999e-07,
+      "loss": 0.5442,
+      "step": 37400
+    },
+    {
+      "epoch": 0.3745,
+      "grad_norm": 4.33549690246582,
+      "learning_rate": 8.5028e-07,
+      "loss": 0.4156,
+      "step": 37450
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 29.745681762695312,
+      "learning_rate": 8.500799999999999e-07,
+      "loss": 0.4352,
+      "step": 37500
+    },
+    {
+      "epoch": 0.3755,
+      "grad_norm": 53.99748611450195,
+      "learning_rate": 8.498799999999999e-07,
+      "loss": 0.5147,
+      "step": 37550
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 11.581042289733887,
+      "learning_rate": 8.4968e-07,
+      "loss": 0.6587,
+      "step": 37600
+    },
+    {
+      "epoch": 0.3765,
+      "grad_norm": 47.363216400146484,
+      "learning_rate": 8.4948e-07,
+      "loss": 0.5124,
+      "step": 37650
+    },
+    {
+      "epoch": 0.377,
+      "grad_norm": 92.71063232421875,
+      "learning_rate": 8.4928e-07,
+      "loss": 0.5673,
+      "step": 37700
+    },
+    {
+      "epoch": 0.3775,
+      "grad_norm": 56.31315994262695,
+      "learning_rate": 8.490799999999999e-07,
+      "loss": 0.5734,
+      "step": 37750
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 66.45987701416016,
+      "learning_rate": 8.488799999999999e-07,
+      "loss": 0.5247,
+      "step": 37800
+    },
+    {
+      "epoch": 0.3785,
+      "grad_norm": 31.38943862915039,
+      "learning_rate": 8.4868e-07,
+      "loss": 0.5605,
+      "step": 37850
+    },
+    {
+      "epoch": 0.379,
+      "grad_norm": 51.36092758178711,
+      "learning_rate": 8.4848e-07,
+      "loss": 0.4554,
+      "step": 37900
+    },
+    {
+      "epoch": 0.3795,
+      "grad_norm": 31.094507217407227,
+      "learning_rate": 8.4828e-07,
+      "loss": 0.6803,
+      "step": 37950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 60.59067916870117,
+      "learning_rate": 8.4808e-07,
+      "loss": 0.6999,
+      "step": 38000
+    },
+    {
+      "epoch": 0.3805,
+      "grad_norm": 11.532386779785156,
+      "learning_rate": 8.478799999999999e-07,
+      "loss": 0.4905,
+      "step": 38050
+    },
+    {
+      "epoch": 0.381,
+      "grad_norm": 77.91910552978516,
+      "learning_rate": 8.476799999999999e-07,
+      "loss": 0.5046,
+      "step": 38100
+    },
+    {
+      "epoch": 0.3815,
+      "grad_norm": 123.77552795410156,
+      "learning_rate": 8.4748e-07,
+      "loss": 0.6066,
+      "step": 38150
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 80.5956802368164,
+      "learning_rate": 8.4728e-07,
+      "loss": 0.5123,
+      "step": 38200
+    },
+    {
+      "epoch": 0.3825,
+      "grad_norm": 18.614927291870117,
+      "learning_rate": 8.4708e-07,
+      "loss": 0.5425,
+      "step": 38250
+    },
+    {
+      "epoch": 0.383,
+      "grad_norm": 40.55034255981445,
+      "learning_rate": 8.4688e-07,
+      "loss": 0.5672,
+      "step": 38300
+    },
+    {
+      "epoch": 0.3835,
+      "grad_norm": 60.523948669433594,
+      "learning_rate": 8.4668e-07,
+      "loss": 0.5903,
+      "step": 38350
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 56.68648910522461,
+      "learning_rate": 8.464799999999999e-07,
+      "loss": 0.5536,
+      "step": 38400
+    },
+    {
+      "epoch": 0.3845,
+      "grad_norm": 63.2244758605957,
+      "learning_rate": 8.4628e-07,
+      "loss": 0.5863,
+      "step": 38450
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 73.95377349853516,
+      "learning_rate": 8.4608e-07,
+      "loss": 0.4515,
+      "step": 38500
+    },
+    {
+      "epoch": 0.3855,
+      "grad_norm": 18.579113006591797,
+      "learning_rate": 8.458799999999999e-07,
+      "loss": 0.6405,
+      "step": 38550
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 40.64234161376953,
+      "learning_rate": 8.4568e-07,
+      "loss": 0.5371,
+      "step": 38600
+    },
+    {
+      "epoch": 0.3865,
+      "grad_norm": 54.03630447387695,
+      "learning_rate": 8.4548e-07,
+      "loss": 0.5378,
+      "step": 38650
+    },
+    {
+      "epoch": 0.387,
+      "grad_norm": 6.5824666023254395,
+      "learning_rate": 8.4528e-07,
+      "loss": 0.4843,
+      "step": 38700
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 8.509979248046875,
+      "learning_rate": 8.4508e-07,
+      "loss": 0.4975,
+      "step": 38750
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 97.18090057373047,
+      "learning_rate": 8.448839999999999e-07,
+      "loss": 0.4974,
+      "step": 38800
+    },
+    {
+      "epoch": 0.3885,
+      "grad_norm": 109.9385757446289,
+      "learning_rate": 8.446839999999999e-07,
+      "loss": 0.5312,
+      "step": 38850
+    },
+    {
+      "epoch": 0.389,
+      "grad_norm": 2.83178448677063,
+      "learning_rate": 8.44484e-07,
+      "loss": 0.4961,
+      "step": 38900
+    },
+    {
+      "epoch": 0.3895,
+      "grad_norm": 1.3601728677749634,
+      "learning_rate": 8.44284e-07,
+      "loss": 0.5074,
+      "step": 38950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 49.6685905456543,
+      "learning_rate": 8.440839999999999e-07,
+      "loss": 0.4328,
+      "step": 39000
+    },
+    {
+      "epoch": 0.3905,
+      "grad_norm": 58.785987854003906,
+      "learning_rate": 8.43884e-07,
+      "loss": 0.5464,
+      "step": 39050
+    },
+    {
+      "epoch": 0.391,
+      "grad_norm": 79.17613220214844,
+      "learning_rate": 8.43688e-07,
+      "loss": 0.5818,
+      "step": 39100
+    },
+    {
+      "epoch": 0.3915,
+      "grad_norm": 30.441478729248047,
+      "learning_rate": 8.434879999999999e-07,
+      "loss": 0.5418,
+      "step": 39150
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 58.006282806396484,
+      "learning_rate": 8.43288e-07,
+      "loss": 0.4618,
+      "step": 39200
+    },
+    {
+      "epoch": 0.3925,
+      "grad_norm": 83.23896026611328,
+      "learning_rate": 8.430879999999999e-07,
+      "loss": 0.6328,
+      "step": 39250
+    },
+    {
+      "epoch": 0.393,
+      "grad_norm": 30.840930938720703,
+      "learning_rate": 8.428879999999999e-07,
+      "loss": 0.52,
+      "step": 39300
+    },
+    {
+      "epoch": 0.3935,
+      "grad_norm": 22.173118591308594,
+      "learning_rate": 8.42688e-07,
+      "loss": 0.4596,
+      "step": 39350
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 123.26498413085938,
+      "learning_rate": 8.42488e-07,
+      "loss": 0.5246,
+      "step": 39400
+    },
+    {
+      "epoch": 0.3945,
+      "grad_norm": 65.1488265991211,
+      "learning_rate": 8.42288e-07,
+      "loss": 0.4235,
+      "step": 39450
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 0.8173670768737793,
+      "learning_rate": 8.420919999999999e-07,
+      "loss": 0.4922,
+      "step": 39500
+    },
+    {
+      "epoch": 0.3955,
+      "grad_norm": 4.30292272567749,
+      "learning_rate": 8.418919999999999e-07,
+      "loss": 0.4174,
+      "step": 39550
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 55.06961441040039,
+      "learning_rate": 8.416919999999999e-07,
+      "loss": 0.5271,
+      "step": 39600
+    },
+    {
+      "epoch": 0.3965,
+      "grad_norm": 7.464923858642578,
+      "learning_rate": 8.41492e-07,
+      "loss": 0.495,
+      "step": 39650
+    },
+    {
+      "epoch": 0.397,
+      "grad_norm": 34.88985824584961,
+      "learning_rate": 8.41292e-07,
+      "loss": 0.5638,
+      "step": 39700
+    },
+    {
+      "epoch": 0.3975,
+      "grad_norm": 36.899593353271484,
+      "learning_rate": 8.410919999999999e-07,
+      "loss": 0.4877,
+      "step": 39750
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 70.25818634033203,
+      "learning_rate": 8.40892e-07,
+      "loss": 0.516,
+      "step": 39800
+    },
+    {
+      "epoch": 0.3985,
+      "grad_norm": 23.857093811035156,
+      "learning_rate": 8.406919999999999e-07,
+      "loss": 0.4775,
+      "step": 39850
+    },
+    {
+      "epoch": 0.399,
+      "grad_norm": 64.42444610595703,
+      "learning_rate": 8.404919999999999e-07,
+      "loss": 0.3856,
+      "step": 39900
+    },
+    {
+      "epoch": 0.3995,
+      "grad_norm": 94.81324005126953,
+      "learning_rate": 8.40292e-07,
+      "loss": 0.603,
+      "step": 39950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 70.74253845214844,
+      "learning_rate": 8.400919999999999e-07,
+      "loss": 0.4607,
+      "step": 40000
+    },
+    {
+      "epoch": 0.4005,
+      "grad_norm": 69.46749877929688,
+      "learning_rate": 8.398919999999999e-07,
+      "loss": 0.5185,
+      "step": 40050
+    },
+    {
+      "epoch": 0.401,
+      "grad_norm": 30.19931983947754,
+      "learning_rate": 8.39692e-07,
+      "loss": 0.5854,
+      "step": 40100
+    },
+    {
+      "epoch": 0.4015,
+      "grad_norm": 80.08529663085938,
+      "learning_rate": 8.39492e-07,
+      "loss": 0.4856,
+      "step": 40150
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 40.47534942626953,
+      "learning_rate": 8.39292e-07,
+      "loss": 0.4393,
+      "step": 40200
+    },
+    {
+      "epoch": 0.4025,
+      "grad_norm": 92.6363296508789,
+      "learning_rate": 8.390919999999999e-07,
+      "loss": 0.4964,
+      "step": 40250
+    },
+    {
+      "epoch": 0.403,
+      "grad_norm": 1.9177128076553345,
+      "learning_rate": 8.388919999999999e-07,
+      "loss": 0.606,
+      "step": 40300
+    },
+    {
+      "epoch": 0.4035,
+      "grad_norm": 9.308789253234863,
+      "learning_rate": 8.38692e-07,
+      "loss": 0.4849,
+      "step": 40350
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 9.60834789276123,
+      "learning_rate": 8.38492e-07,
+      "loss": 0.5118,
+      "step": 40400
+    },
+    {
+      "epoch": 0.4045,
+      "grad_norm": 70.33528137207031,
+      "learning_rate": 8.38292e-07,
+      "loss": 0.5834,
+      "step": 40450
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 93.31282806396484,
+      "learning_rate": 8.38092e-07,
+      "loss": 0.5711,
+      "step": 40500
+    },
+    {
+      "epoch": 0.4055,
+      "grad_norm": 56.34794998168945,
+      "learning_rate": 8.378919999999999e-07,
+      "loss": 0.5346,
+      "step": 40550
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 53.3722038269043,
+      "learning_rate": 8.376919999999999e-07,
+      "loss": 0.5983,
+      "step": 40600
+    },
+    {
+      "epoch": 0.4065,
+      "grad_norm": 79.872314453125,
+      "learning_rate": 8.37492e-07,
+      "loss": 0.6024,
+      "step": 40650
+    },
+    {
+      "epoch": 0.407,
+      "grad_norm": 43.69386291503906,
+      "learning_rate": 8.37292e-07,
+      "loss": 0.6604,
+      "step": 40700
+    },
+    {
+      "epoch": 0.4075,
+      "grad_norm": 82.44725799560547,
+      "learning_rate": 8.370919999999999e-07,
+      "loss": 0.5791,
+      "step": 40750
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 28.565074920654297,
+      "learning_rate": 8.36892e-07,
+      "loss": 0.4985,
+      "step": 40800
+    },
+    {
+      "epoch": 0.4085,
+      "grad_norm": 108.46695709228516,
+      "learning_rate": 8.36692e-07,
+      "loss": 0.559,
+      "step": 40850
+    },
+    {
+      "epoch": 0.409,
+      "grad_norm": 68.24771881103516,
+      "learning_rate": 8.364919999999999e-07,
+      "loss": 0.6729,
+      "step": 40900
+    },
+    {
+      "epoch": 0.4095,
+      "grad_norm": 106.6298599243164,
+      "learning_rate": 8.36292e-07,
+      "loss": 0.6222,
+      "step": 40950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.868021011352539,
+      "learning_rate": 8.360919999999999e-07,
+      "loss": 0.4931,
+      "step": 41000
+    },
+    {
+      "epoch": 0.4105,
+      "grad_norm": 77.19261169433594,
+      "learning_rate": 8.358919999999999e-07,
+      "loss": 0.5437,
+      "step": 41050
+    },
+    {
+      "epoch": 0.411,
+      "grad_norm": 52.96466064453125,
+      "learning_rate": 8.35692e-07,
+      "loss": 0.5333,
+      "step": 41100
+    },
+    {
+      "epoch": 0.4115,
+      "grad_norm": 47.27060317993164,
+      "learning_rate": 8.35492e-07,
+      "loss": 0.6691,
+      "step": 41150
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 65.95133209228516,
+      "learning_rate": 8.35292e-07,
+      "loss": 0.506,
+      "step": 41200
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 124.53824615478516,
+      "learning_rate": 8.350919999999999e-07,
+      "loss": 0.5634,
+      "step": 41250
+    },
+    {
+      "epoch": 0.413,
+      "grad_norm": 59.79798126220703,
+      "learning_rate": 8.348919999999999e-07,
+      "loss": 0.6084,
+      "step": 41300
+    },
+    {
+      "epoch": 0.4135,
+      "grad_norm": 17.437131881713867,
+      "learning_rate": 8.346919999999999e-07,
+      "loss": 0.4448,
+      "step": 41350
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 0.26926401257514954,
+      "learning_rate": 8.34492e-07,
+      "loss": 0.5695,
+      "step": 41400
+    },
+    {
+      "epoch": 0.4145,
+      "grad_norm": 25.987173080444336,
+      "learning_rate": 8.34292e-07,
+      "loss": 0.4574,
+      "step": 41450
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 77.89625549316406,
+      "learning_rate": 8.340919999999999e-07,
+      "loss": 0.5357,
+      "step": 41500
+    },
+    {
+      "epoch": 0.4155,
+      "grad_norm": 37.63655471801758,
+      "learning_rate": 8.33892e-07,
+      "loss": 0.5492,
+      "step": 41550
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 103.72399139404297,
+      "learning_rate": 8.33696e-07,
+      "loss": 0.4425,
+      "step": 41600
+    },
+    {
+      "epoch": 0.4165,
+      "grad_norm": 73.7927017211914,
+      "learning_rate": 8.334959999999999e-07,
+      "loss": 0.54,
+      "step": 41650
+    },
+    {
+      "epoch": 0.417,
+      "grad_norm": 43.02588653564453,
+      "learning_rate": 8.33296e-07,
+      "loss": 0.492,
+      "step": 41700
+    },
+    {
+      "epoch": 0.4175,
+      "grad_norm": 71.02433776855469,
+      "learning_rate": 8.330959999999999e-07,
+      "loss": 0.5462,
+      "step": 41750
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 115.1431884765625,
+      "learning_rate": 8.328959999999999e-07,
+      "loss": 0.5912,
+      "step": 41800
+    },
+    {
+      "epoch": 0.4185,
+      "grad_norm": 96.75558471679688,
+      "learning_rate": 8.32696e-07,
+      "loss": 0.5876,
+      "step": 41850
+    },
+    {
+      "epoch": 0.419,
+      "grad_norm": 36.395286560058594,
+      "learning_rate": 8.32496e-07,
+      "loss": 0.4394,
+      "step": 41900
+    },
+    {
+      "epoch": 0.4195,
+      "grad_norm": 168.5052947998047,
+      "learning_rate": 8.32296e-07,
+      "loss": 0.5484,
+      "step": 41950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 60.965389251708984,
+      "learning_rate": 8.320959999999999e-07,
+      "loss": 0.4683,
+      "step": 42000
+    },
+    {
+      "epoch": 0.4205,
+      "grad_norm": 118.09526062011719,
+      "learning_rate": 8.318959999999999e-07,
+      "loss": 0.5096,
+      "step": 42050
+    },
+    {
+      "epoch": 0.421,
+      "grad_norm": 91.55111694335938,
+      "learning_rate": 8.316959999999999e-07,
+      "loss": 0.5228,
+      "step": 42100
+    },
+    {
+      "epoch": 0.4215,
+      "grad_norm": 24.228757858276367,
+      "learning_rate": 8.31496e-07,
+      "loss": 0.5189,
+      "step": 42150
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 5.371842861175537,
+      "learning_rate": 8.31296e-07,
+      "loss": 0.4865,
+      "step": 42200
+    },
+    {
+      "epoch": 0.4225,
+      "grad_norm": 74.1497802734375,
+      "learning_rate": 8.310959999999999e-07,
+      "loss": 0.4502,
+      "step": 42250
+    },
+    {
+      "epoch": 0.423,
+      "grad_norm": 3.059469223022461,
+      "learning_rate": 8.30896e-07,
+      "loss": 0.4839,
+      "step": 42300
+    },
+    {
+      "epoch": 0.4235,
+      "grad_norm": 14.30676555633545,
+      "learning_rate": 8.306959999999999e-07,
+      "loss": 0.4863,
+      "step": 42350
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 0.32036876678466797,
+      "learning_rate": 8.30496e-07,
+      "loss": 0.4859,
+      "step": 42400
+    },
+    {
+      "epoch": 0.4245,
+      "grad_norm": 36.87676239013672,
+      "learning_rate": 8.30296e-07,
+      "loss": 0.3993,
+      "step": 42450
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 53.69343566894531,
+      "learning_rate": 8.300959999999999e-07,
+      "loss": 0.564,
+      "step": 42500
+    },
+    {
+      "epoch": 0.4255,
+      "grad_norm": 49.399715423583984,
+      "learning_rate": 8.29896e-07,
+      "loss": 0.537,
+      "step": 42550
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 42.28934860229492,
+      "learning_rate": 8.29696e-07,
+      "loss": 0.4264,
+      "step": 42600
+    },
+    {
+      "epoch": 0.4265,
+      "grad_norm": 14.933109283447266,
+      "learning_rate": 8.29496e-07,
+      "loss": 0.5168,
+      "step": 42650
+    },
+    {
+      "epoch": 0.427,
+      "grad_norm": 69.50049591064453,
+      "learning_rate": 8.29296e-07,
+      "loss": 0.5216,
+      "step": 42700
+    },
+    {
+      "epoch": 0.4275,
+      "grad_norm": 6.045228958129883,
+      "learning_rate": 8.29096e-07,
+      "loss": 0.6186,
+      "step": 42750
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 17.617143630981445,
+      "learning_rate": 8.288959999999999e-07,
+      "loss": 0.5476,
+      "step": 42800
+    },
+    {
+      "epoch": 0.4285,
+      "grad_norm": 21.75743293762207,
+      "learning_rate": 8.28696e-07,
+      "loss": 0.4667,
+      "step": 42850
+    },
+    {
+      "epoch": 0.429,
+      "grad_norm": 85.2100830078125,
+      "learning_rate": 8.28496e-07,
+      "loss": 0.6098,
+      "step": 42900
+    },
+    {
+      "epoch": 0.4295,
+      "grad_norm": 10.500739097595215,
+      "learning_rate": 8.28296e-07,
+      "loss": 0.434,
+      "step": 42950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 42.483436584472656,
+      "learning_rate": 8.280960000000001e-07,
+      "loss": 0.4656,
+      "step": 43000
+    },
+    {
+      "epoch": 0.4305,
+      "grad_norm": 21.001169204711914,
+      "learning_rate": 8.278959999999999e-07,
+      "loss": 0.5221,
+      "step": 43050
+    },
+    {
+      "epoch": 0.431,
+      "grad_norm": 17.075965881347656,
+      "learning_rate": 8.276959999999999e-07,
+      "loss": 0.4657,
+      "step": 43100
+    },
+    {
+      "epoch": 0.4315,
+      "grad_norm": 15.272270202636719,
+      "learning_rate": 8.27496e-07,
+      "loss": 0.4914,
+      "step": 43150
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 51.080135345458984,
+      "learning_rate": 8.27296e-07,
+      "loss": 0.689,
+      "step": 43200
+    },
+    {
+      "epoch": 0.4325,
+      "grad_norm": 99.12468719482422,
+      "learning_rate": 8.27096e-07,
+      "loss": 0.4991,
+      "step": 43250
+    },
+    {
+      "epoch": 0.433,
+      "grad_norm": 69.47232055664062,
+      "learning_rate": 8.26896e-07,
+      "loss": 0.4888,
+      "step": 43300
+    },
+    {
+      "epoch": 0.4335,
+      "grad_norm": 71.0149917602539,
+      "learning_rate": 8.26696e-07,
+      "loss": 0.625,
+      "step": 43350
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 47.15895080566406,
+      "learning_rate": 8.264959999999999e-07,
+      "loss": 0.5997,
+      "step": 43400
+    },
+    {
+      "epoch": 0.4345,
+      "grad_norm": 74.67218017578125,
+      "learning_rate": 8.26296e-07,
+      "loss": 0.5972,
+      "step": 43450
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 53.72488784790039,
+      "learning_rate": 8.26096e-07,
+      "loss": 0.5028,
+      "step": 43500
+    },
+    {
+      "epoch": 0.4355,
+      "grad_norm": 14.094425201416016,
+      "learning_rate": 8.258959999999999e-07,
+      "loss": 0.5297,
+      "step": 43550
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 12.79079818725586,
+      "learning_rate": 8.25696e-07,
+      "loss": 0.4202,
+      "step": 43600
+    },
+    {
+      "epoch": 0.4365,
+      "grad_norm": 28.30698013305664,
+      "learning_rate": 8.25496e-07,
+      "loss": 0.4182,
+      "step": 43650
+    },
+    {
+      "epoch": 0.437,
+      "grad_norm": 91.74481964111328,
+      "learning_rate": 8.25296e-07,
+      "loss": 0.4885,
+      "step": 43700
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 13.599432945251465,
+      "learning_rate": 8.25096e-07,
+      "loss": 0.6304,
+      "step": 43750
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 3.0752172470092773,
+      "learning_rate": 8.248959999999999e-07,
+      "loss": 0.449,
+      "step": 43800
+    },
+    {
+      "epoch": 0.4385,
+      "grad_norm": 74.79292297363281,
+      "learning_rate": 8.246959999999999e-07,
+      "loss": 0.6306,
+      "step": 43850
+    },
+    {
+      "epoch": 0.439,
+      "grad_norm": 43.5394401550293,
+      "learning_rate": 8.24496e-07,
+      "loss": 0.6897,
+      "step": 43900
+    },
+    {
+      "epoch": 0.4395,
+      "grad_norm": 25.60984230041504,
+      "learning_rate": 8.24296e-07,
+      "loss": 0.5346,
+      "step": 43950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 91.31788635253906,
+      "learning_rate": 8.24096e-07,
+      "loss": 0.6347,
+      "step": 44000
+    },
+    {
+      "epoch": 0.4405,
+      "grad_norm": 90.00495147705078,
+      "learning_rate": 8.23896e-07,
+      "loss": 0.5585,
+      "step": 44050
+    },
+    {
+      "epoch": 0.441,
+      "grad_norm": 78.50735473632812,
+      "learning_rate": 8.236959999999999e-07,
+      "loss": 0.5065,
+      "step": 44100
+    },
+    {
+      "epoch": 0.4415,
+      "grad_norm": 61.63063049316406,
+      "learning_rate": 8.234959999999999e-07,
+      "loss": 0.3718,
+      "step": 44150
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 64.05858612060547,
+      "learning_rate": 8.23296e-07,
+      "loss": 0.5225,
+      "step": 44200
+    },
+    {
+      "epoch": 0.4425,
+      "grad_norm": 46.02941131591797,
+      "learning_rate": 8.23096e-07,
+      "loss": 0.4429,
+      "step": 44250
+    },
+    {
+      "epoch": 0.443,
+      "grad_norm": 19.115480422973633,
+      "learning_rate": 8.228959999999999e-07,
+      "loss": 0.6595,
+      "step": 44300
+    },
+    {
+      "epoch": 0.4435,
+      "grad_norm": 5.615220069885254,
+      "learning_rate": 8.22696e-07,
+      "loss": 0.3437,
+      "step": 44350
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 53.1685791015625,
+      "learning_rate": 8.22496e-07,
+      "loss": 0.5388,
+      "step": 44400
+    },
+    {
+      "epoch": 0.4445,
+      "grad_norm": 35.044410705566406,
+      "learning_rate": 8.222959999999999e-07,
+      "loss": 0.6554,
+      "step": 44450
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 98.26656341552734,
+      "learning_rate": 8.22096e-07,
+      "loss": 0.5233,
+      "step": 44500
+    },
+    {
+      "epoch": 0.4455,
+      "grad_norm": 109.79158782958984,
+      "learning_rate": 8.218959999999999e-07,
+      "loss": 0.5021,
+      "step": 44550
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 59.18463134765625,
+      "learning_rate": 8.216959999999999e-07,
+      "loss": 0.5612,
+      "step": 44600
+    },
+    {
+      "epoch": 0.4465,
+      "grad_norm": 89.63908386230469,
+      "learning_rate": 8.21496e-07,
+      "loss": 0.4865,
+      "step": 44650
+    },
+    {
+      "epoch": 0.447,
+      "grad_norm": 75.61219024658203,
+      "learning_rate": 8.21296e-07,
+      "loss": 0.4486,
+      "step": 44700
+    },
+    {
+      "epoch": 0.4475,
+      "grad_norm": 4.093531131744385,
+      "learning_rate": 8.210960000000001e-07,
+      "loss": 0.6416,
+      "step": 44750
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 37.043434143066406,
+      "learning_rate": 8.208959999999999e-07,
+      "loss": 0.6513,
+      "step": 44800
+    },
+    {
+      "epoch": 0.4485,
+      "grad_norm": 51.53557586669922,
+      "learning_rate": 8.206959999999999e-07,
+      "loss": 0.4894,
+      "step": 44850
+    },
+    {
+      "epoch": 0.449,
+      "grad_norm": 69.94869232177734,
+      "learning_rate": 8.20496e-07,
+      "loss": 0.4744,
+      "step": 44900
+    },
+    {
+      "epoch": 0.4495,
+      "grad_norm": 76.57780456542969,
+      "learning_rate": 8.20296e-07,
+      "loss": 0.5147,
+      "step": 44950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 82.7248764038086,
+      "learning_rate": 8.20096e-07,
+      "loss": 0.6523,
+      "step": 45000
+    },
+    {
+      "epoch": 0.4505,
+      "grad_norm": 23.24110984802246,
+      "learning_rate": 8.19896e-07,
+      "loss": 0.5502,
+      "step": 45050
+    },
+    {
+      "epoch": 0.451,
+      "grad_norm": 60.64106750488281,
+      "learning_rate": 8.19696e-07,
+      "loss": 0.4758,
+      "step": 45100
+    },
+    {
+      "epoch": 0.4515,
+      "grad_norm": 35.0093879699707,
+      "learning_rate": 8.194959999999999e-07,
+      "loss": 0.544,
+      "step": 45150
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 27.198448181152344,
+      "learning_rate": 8.19296e-07,
+      "loss": 0.5395,
+      "step": 45200
+    },
+    {
+      "epoch": 0.4525,
+      "grad_norm": 51.811222076416016,
+      "learning_rate": 8.19096e-07,
+      "loss": 0.5858,
+      "step": 45250
+    },
+    {
+      "epoch": 0.453,
+      "grad_norm": 59.12669372558594,
+      "learning_rate": 8.188959999999999e-07,
+      "loss": 0.4579,
+      "step": 45300
+    },
+    {
+      "epoch": 0.4535,
+      "grad_norm": 35.265716552734375,
+      "learning_rate": 8.18696e-07,
+      "loss": 0.53,
+      "step": 45350
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 38.40176010131836,
+      "learning_rate": 8.18496e-07,
+      "loss": 0.5556,
+      "step": 45400
+    },
+    {
+      "epoch": 0.4545,
+      "grad_norm": 6.6817240715026855,
+      "learning_rate": 8.18296e-07,
+      "loss": 0.4169,
+      "step": 45450
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 77.1117935180664,
+      "learning_rate": 8.18096e-07,
+      "loss": 0.5975,
+      "step": 45500
+    },
+    {
+      "epoch": 0.4555,
+      "grad_norm": 32.58802032470703,
+      "learning_rate": 8.178959999999999e-07,
+      "loss": 0.4934,
+      "step": 45550
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 80.36177062988281,
+      "learning_rate": 8.176959999999999e-07,
+      "loss": 0.4513,
+      "step": 45600
+    },
+    {
+      "epoch": 0.4565,
+      "grad_norm": 38.63931655883789,
+      "learning_rate": 8.17496e-07,
+      "loss": 0.4297,
+      "step": 45650
+    },
+    {
+      "epoch": 0.457,
+      "grad_norm": 30.655248641967773,
+      "learning_rate": 8.17296e-07,
+      "loss": 0.6114,
+      "step": 45700
+    },
+    {
+      "epoch": 0.4575,
+      "grad_norm": 122.92070007324219,
+      "learning_rate": 8.17096e-07,
+      "loss": 0.5557,
+      "step": 45750
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 63.41127395629883,
+      "learning_rate": 8.16896e-07,
+      "loss": 0.6203,
+      "step": 45800
+    },
+    {
+      "epoch": 0.4585,
+      "grad_norm": 10.74167537689209,
+      "learning_rate": 8.166959999999999e-07,
+      "loss": 0.4382,
+      "step": 45850
+    },
+    {
+      "epoch": 0.459,
+      "grad_norm": 70.49405670166016,
+      "learning_rate": 8.164959999999999e-07,
+      "loss": 0.4747,
+      "step": 45900
+    },
+    {
+      "epoch": 0.4595,
+      "grad_norm": 3.7027602195739746,
+      "learning_rate": 8.16296e-07,
+      "loss": 0.6623,
+      "step": 45950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 25.70686149597168,
+      "learning_rate": 8.16096e-07,
+      "loss": 0.494,
+      "step": 46000
+    },
+    {
+      "epoch": 0.4605,
+      "grad_norm": 32.90084457397461,
+      "learning_rate": 8.158959999999999e-07,
+      "loss": 0.4998,
+      "step": 46050
+    },
+    {
+      "epoch": 0.461,
+      "grad_norm": 14.209221839904785,
+      "learning_rate": 8.15696e-07,
+      "loss": 0.4964,
+      "step": 46100
+    },
+    {
+      "epoch": 0.4615,
+      "grad_norm": 0.3458581268787384,
+      "learning_rate": 8.15496e-07,
+      "loss": 0.5275,
+      "step": 46150
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 66.87709045410156,
+      "learning_rate": 8.152959999999999e-07,
+      "loss": 0.5533,
+      "step": 46200
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 53.677860260009766,
+      "learning_rate": 8.15096e-07,
+      "loss": 0.5109,
+      "step": 46250
+    },
+    {
+      "epoch": 0.463,
+      "grad_norm": 43.0468864440918,
+      "learning_rate": 8.148959999999999e-07,
+      "loss": 0.4664,
+      "step": 46300
+    },
+    {
+      "epoch": 0.4635,
+      "grad_norm": 70.673095703125,
+      "learning_rate": 8.146959999999999e-07,
+      "loss": 0.6125,
+      "step": 46350
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 71.34368133544922,
+      "learning_rate": 8.14496e-07,
+      "loss": 0.4832,
+      "step": 46400
+    },
+    {
+      "epoch": 0.4645,
+      "grad_norm": 198.13221740722656,
+      "learning_rate": 8.14296e-07,
+      "loss": 0.6594,
+      "step": 46450
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 35.54419708251953,
+      "learning_rate": 8.14096e-07,
+      "loss": 0.5413,
+      "step": 46500
+    },
+    {
+      "epoch": 0.4655,
+      "grad_norm": 78.56597900390625,
+      "learning_rate": 8.138959999999999e-07,
+      "loss": 0.6681,
+      "step": 46550
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 22.999269485473633,
+      "learning_rate": 8.136959999999999e-07,
+      "loss": 0.4987,
+      "step": 46600
+    },
+    {
+      "epoch": 0.4665,
+      "grad_norm": 1.7849301099777222,
+      "learning_rate": 8.134959999999999e-07,
+      "loss": 0.4031,
+      "step": 46650
+    },
+    {
+      "epoch": 0.467,
+      "grad_norm": 27.667150497436523,
+      "learning_rate": 8.13296e-07,
+      "loss": 0.5426,
+      "step": 46700
+    },
+    {
+      "epoch": 0.4675,
+      "grad_norm": 33.66166687011719,
+      "learning_rate": 8.13096e-07,
+      "loss": 0.4668,
+      "step": 46750
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 28.29575538635254,
+      "learning_rate": 8.128959999999999e-07,
+      "loss": 0.5348,
+      "step": 46800
+    },
+    {
+      "epoch": 0.4685,
+      "grad_norm": 48.07258605957031,
+      "learning_rate": 8.12696e-07,
+      "loss": 0.7829,
+      "step": 46850
+    },
+    {
+      "epoch": 0.469,
+      "grad_norm": 42.36520004272461,
+      "learning_rate": 8.124959999999999e-07,
+      "loss": 0.6122,
+      "step": 46900
+    },
+    {
+      "epoch": 0.4695,
+      "grad_norm": 106.2151107788086,
+      "learning_rate": 8.122959999999999e-07,
+      "loss": 0.621,
+      "step": 46950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 76.08829498291016,
+      "learning_rate": 8.12096e-07,
+      "loss": 0.6716,
+      "step": 47000
+    },
+    {
+      "epoch": 0.4705,
+      "grad_norm": 31.28024673461914,
+      "learning_rate": 8.118959999999999e-07,
+      "loss": 0.4718,
+      "step": 47050
+    },
+    {
+      "epoch": 0.471,
+      "grad_norm": 55.14548873901367,
+      "learning_rate": 8.116959999999999e-07,
+      "loss": 0.6012,
+      "step": 47100
+    },
+    {
+      "epoch": 0.4715,
+      "grad_norm": 8.463032722473145,
+      "learning_rate": 8.11496e-07,
+      "loss": 0.4563,
+      "step": 47150
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 74.97944641113281,
+      "learning_rate": 8.11296e-07,
+      "loss": 0.568,
+      "step": 47200
+    },
+    {
+      "epoch": 0.4725,
+      "grad_norm": 1.081640601158142,
+      "learning_rate": 8.11096e-07,
+      "loss": 0.6412,
+      "step": 47250
+    },
+    {
+      "epoch": 0.473,
+      "grad_norm": 77.51432800292969,
+      "learning_rate": 8.108959999999999e-07,
+      "loss": 0.4582,
+      "step": 47300
+    },
+    {
+      "epoch": 0.4735,
+      "grad_norm": 45.42144012451172,
+      "learning_rate": 8.106959999999999e-07,
+      "loss": 0.4746,
+      "step": 47350
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 0.2983810603618622,
+      "learning_rate": 8.105e-07,
+      "loss": 0.6629,
+      "step": 47400
+    },
+    {
+      "epoch": 0.4745,
+      "grad_norm": 5.902935981750488,
+      "learning_rate": 8.103e-07,
+      "loss": 0.5522,
+      "step": 47450
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 0.9811944961547852,
+      "learning_rate": 8.101e-07,
+      "loss": 0.4998,
+      "step": 47500
+    },
+    {
+      "epoch": 0.4755,
+      "grad_norm": 21.928115844726562,
+      "learning_rate": 8.099e-07,
+      "loss": 0.6313,
+      "step": 47550
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 51.325443267822266,
+      "learning_rate": 8.097e-07,
+      "loss": 0.5051,
+      "step": 47600
+    },
+    {
+      "epoch": 0.4765,
+      "grad_norm": 10.640209197998047,
+      "learning_rate": 8.094999999999999e-07,
+      "loss": 0.4917,
+      "step": 47650
+    },
+    {
+      "epoch": 0.477,
+      "grad_norm": 90.27954864501953,
+      "learning_rate": 8.093e-07,
+      "loss": 0.4371,
+      "step": 47700
+    },
+    {
+      "epoch": 0.4775,
+      "grad_norm": 78.23124694824219,
+      "learning_rate": 8.091e-07,
+      "loss": 0.5052,
+      "step": 47750
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 3.3375375270843506,
+      "learning_rate": 8.088999999999999e-07,
+      "loss": 0.5095,
+      "step": 47800
+    },
+    {
+      "epoch": 0.4785,
+      "grad_norm": 17.96718406677246,
+      "learning_rate": 8.087e-07,
+      "loss": 0.4613,
+      "step": 47850
+    },
+    {
+      "epoch": 0.479,
+      "grad_norm": 86.63037872314453,
+      "learning_rate": 8.085e-07,
+      "loss": 0.5274,
+      "step": 47900
+    },
+    {
+      "epoch": 0.4795,
+      "grad_norm": 15.691425323486328,
+      "learning_rate": 8.083e-07,
+      "loss": 0.5586,
+      "step": 47950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 95.55493927001953,
+      "learning_rate": 8.081e-07,
+      "loss": 0.5387,
+      "step": 48000
+    },
+    {
+      "epoch": 0.4805,
+      "grad_norm": 86.83247375488281,
+      "learning_rate": 8.078999999999999e-07,
+      "loss": 0.5033,
+      "step": 48050
+    },
+    {
+      "epoch": 0.481,
+      "grad_norm": 24.38957405090332,
+      "learning_rate": 8.076999999999999e-07,
+      "loss": 0.7073,
+      "step": 48100
+    },
+    {
+      "epoch": 0.4815,
+      "grad_norm": 49.8204231262207,
+      "learning_rate": 8.075e-07,
+      "loss": 0.4825,
+      "step": 48150
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 81.01677703857422,
+      "learning_rate": 8.07304e-07,
+      "loss": 0.6172,
+      "step": 48200
+    },
+    {
+      "epoch": 0.4825,
+      "grad_norm": 9.030482292175293,
+      "learning_rate": 8.07104e-07,
+      "loss": 0.5569,
+      "step": 48250
+    },
+    {
+      "epoch": 0.483,
+      "grad_norm": 94.63253021240234,
+      "learning_rate": 8.06904e-07,
+      "loss": 0.5637,
+      "step": 48300
+    },
+    {
+      "epoch": 0.4835,
+      "grad_norm": 92.56613159179688,
+      "learning_rate": 8.067039999999999e-07,
+      "loss": 0.6043,
+      "step": 48350
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 21.853906631469727,
+      "learning_rate": 8.065039999999999e-07,
+      "loss": 0.4438,
+      "step": 48400
+    },
+    {
+      "epoch": 0.4845,
+      "grad_norm": 22.982881546020508,
+      "learning_rate": 8.06304e-07,
+      "loss": 0.498,
+      "step": 48450
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 57.36958312988281,
+      "learning_rate": 8.06104e-07,
+      "loss": 0.5241,
+      "step": 48500
+    },
+    {
+      "epoch": 0.4855,
+      "grad_norm": 21.14190673828125,
+      "learning_rate": 8.059039999999999e-07,
+      "loss": 0.6204,
+      "step": 48550
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 52.20134353637695,
+      "learning_rate": 8.05704e-07,
+      "loss": 0.5012,
+      "step": 48600
+    },
+    {
+      "epoch": 0.4865,
+      "grad_norm": 1.349229335784912,
+      "learning_rate": 8.05504e-07,
+      "loss": 0.5246,
+      "step": 48650
+    },
+    {
+      "epoch": 0.487,
+      "grad_norm": 66.7760009765625,
+      "learning_rate": 8.053039999999999e-07,
+      "loss": 0.5898,
+      "step": 48700
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 20.221620559692383,
+      "learning_rate": 8.05104e-07,
+      "loss": 0.449,
+      "step": 48750
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 36.600914001464844,
+      "learning_rate": 8.049039999999999e-07,
+      "loss": 0.4766,
+      "step": 48800
+    },
+    {
+      "epoch": 0.4885,
+      "grad_norm": 65.7702865600586,
+      "learning_rate": 8.047039999999999e-07,
+      "loss": 0.66,
+      "step": 48850
+    },
+    {
+      "epoch": 0.489,
+      "grad_norm": 41.316524505615234,
+      "learning_rate": 8.04504e-07,
+      "loss": 0.549,
+      "step": 48900
+    },
+    {
+      "epoch": 0.4895,
+      "grad_norm": 30.280691146850586,
+      "learning_rate": 8.04304e-07,
+      "loss": 0.5504,
+      "step": 48950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 34.455535888671875,
+      "learning_rate": 8.04104e-07,
+      "loss": 0.502,
+      "step": 49000
+    },
+    {
+      "epoch": 0.4905,
+      "grad_norm": 57.2782096862793,
+      "learning_rate": 8.039039999999999e-07,
+      "loss": 0.4723,
+      "step": 49050
+    },
+    {
+      "epoch": 0.491,
+      "grad_norm": 29.322336196899414,
+      "learning_rate": 8.037039999999999e-07,
+      "loss": 0.5335,
+      "step": 49100
+    },
+    {
+      "epoch": 0.4915,
+      "grad_norm": 88.96637725830078,
+      "learning_rate": 8.035039999999999e-07,
+      "loss": 0.5582,
+      "step": 49150
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 59.695770263671875,
+      "learning_rate": 8.03304e-07,
+      "loss": 0.5627,
+      "step": 49200
+    },
+    {
+      "epoch": 0.4925,
+      "grad_norm": 96.95304107666016,
+      "learning_rate": 8.03104e-07,
+      "loss": 0.5773,
+      "step": 49250
+    },
+    {
+      "epoch": 0.493,
+      "grad_norm": 76.24807739257812,
+      "learning_rate": 8.029039999999999e-07,
+      "loss": 0.5979,
+      "step": 49300
+    },
+    {
+      "epoch": 0.4935,
+      "grad_norm": 5.769176483154297,
+      "learning_rate": 8.02704e-07,
+      "loss": 0.4968,
+      "step": 49350
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 61.573577880859375,
+      "learning_rate": 8.025039999999999e-07,
+      "loss": 0.5445,
+      "step": 49400
+    },
+    {
+      "epoch": 0.4945,
+      "grad_norm": 66.46489715576172,
+      "learning_rate": 8.023039999999999e-07,
+      "loss": 0.467,
+      "step": 49450
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 84.0884780883789,
+      "learning_rate": 8.02104e-07,
+      "loss": 0.5543,
+      "step": 49500
+    },
+    {
+      "epoch": 0.4955,
+      "grad_norm": 73.09436798095703,
+      "learning_rate": 8.019039999999999e-07,
+      "loss": 0.5498,
+      "step": 49550
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 17.9005126953125,
+      "learning_rate": 8.017039999999999e-07,
+      "loss": 0.4723,
+      "step": 49600
+    },
+    {
+      "epoch": 0.4965,
+      "grad_norm": 50.205223083496094,
+      "learning_rate": 8.01504e-07,
+      "loss": 0.5561,
+      "step": 49650
+    },
+    {
+      "epoch": 0.497,
+      "grad_norm": 52.76295471191406,
+      "learning_rate": 8.01304e-07,
+      "loss": 0.4949,
+      "step": 49700
+    },
+    {
+      "epoch": 0.4975,
+      "grad_norm": 62.78417205810547,
+      "learning_rate": 8.01104e-07,
+      "loss": 0.56,
+      "step": 49750
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 6.838378429412842,
+      "learning_rate": 8.009039999999999e-07,
+      "loss": 0.6055,
+      "step": 49800
+    },
+    {
+      "epoch": 0.4985,
+      "grad_norm": 18.722816467285156,
+      "learning_rate": 8.007039999999999e-07,
+      "loss": 0.5358,
+      "step": 49850
+    },
+    {
+      "epoch": 0.499,
+      "grad_norm": 72.58301544189453,
+      "learning_rate": 8.00504e-07,
+      "loss": 0.401,
+      "step": 49900
+    },
+    {
+      "epoch": 0.4995,
+      "grad_norm": 37.88386535644531,
+      "learning_rate": 8.00304e-07,
+      "loss": 0.5257,
+      "step": 49950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 72.8793716430664,
+      "learning_rate": 8.00104e-07,
+      "loss": 0.5824,
+      "step": 50000
+    },
+    {
+      "epoch": 0.5005,
+      "grad_norm": 64.095458984375,
+      "learning_rate": 7.99904e-07,
+      "loss": 0.5659,
+      "step": 50050
+    },
+    {
+      "epoch": 0.501,
+      "grad_norm": 25.19496726989746,
+      "learning_rate": 7.997039999999999e-07,
+      "loss": 0.4465,
+      "step": 50100
+    },
+    {
+      "epoch": 0.5015,
+      "grad_norm": 112.08228302001953,
+      "learning_rate": 7.995039999999999e-07,
+      "loss": 0.6215,
+      "step": 50150
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 3.157916307449341,
+      "learning_rate": 7.99304e-07,
+      "loss": 0.6333,
+      "step": 50200
+    },
+    {
+      "epoch": 0.5025,
+      "grad_norm": 76.75867462158203,
+      "learning_rate": 7.99104e-07,
+      "loss": 0.483,
+      "step": 50250
+    },
+    {
+      "epoch": 0.503,
+      "grad_norm": 43.23719024658203,
+      "learning_rate": 7.989039999999999e-07,
+      "loss": 0.5854,
+      "step": 50300
+    },
+    {
+      "epoch": 0.5035,
+      "grad_norm": 68.20858001708984,
+      "learning_rate": 7.98708e-07,
+      "loss": 0.6767,
+      "step": 50350
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 88.28502655029297,
+      "learning_rate": 7.98508e-07,
+      "loss": 0.7013,
+      "step": 50400
+    },
+    {
+      "epoch": 0.5045,
+      "grad_norm": 36.20711898803711,
+      "learning_rate": 7.98308e-07,
+      "loss": 0.3669,
+      "step": 50450
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 36.513954162597656,
+      "learning_rate": 7.98108e-07,
+      "loss": 0.4585,
+      "step": 50500
+    },
+    {
+      "epoch": 0.5055,
+      "grad_norm": 58.205039978027344,
+      "learning_rate": 7.979079999999999e-07,
+      "loss": 0.4132,
+      "step": 50550
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 79.69970703125,
+      "learning_rate": 7.977079999999999e-07,
+      "loss": 0.5174,
+      "step": 50600
+    },
+    {
+      "epoch": 0.5065,
+      "grad_norm": 45.726173400878906,
+      "learning_rate": 7.97508e-07,
+      "loss": 0.5019,
+      "step": 50650
+    },
+    {
+      "epoch": 0.507,
+      "grad_norm": 88.03544616699219,
+      "learning_rate": 7.97308e-07,
+      "loss": 0.4937,
+      "step": 50700
+    },
+    {
+      "epoch": 0.5075,
+      "grad_norm": 47.83407211303711,
+      "learning_rate": 7.97108e-07,
+      "loss": 0.5023,
+      "step": 50750
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 28.184261322021484,
+      "learning_rate": 7.96908e-07,
+      "loss": 0.4929,
+      "step": 50800
+    },
+    {
+      "epoch": 0.5085,
+      "grad_norm": 97.35902404785156,
+      "learning_rate": 7.967079999999999e-07,
+      "loss": 0.4831,
+      "step": 50850
+    },
+    {
+      "epoch": 0.509,
+      "grad_norm": 45.97766876220703,
+      "learning_rate": 7.965079999999999e-07,
+      "loss": 0.5077,
+      "step": 50900
+    },
+    {
+      "epoch": 0.5095,
+      "grad_norm": 126.22740173339844,
+      "learning_rate": 7.96308e-07,
+      "loss": 0.7102,
+      "step": 50950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 47.19316101074219,
+      "learning_rate": 7.96108e-07,
+      "loss": 0.4501,
+      "step": 51000
+    },
+    {
+      "epoch": 0.5105,
+      "grad_norm": 7.235782146453857,
+      "learning_rate": 7.959079999999999e-07,
+      "loss": 0.4165,
+      "step": 51050
+    },
+    {
+      "epoch": 0.511,
+      "grad_norm": 1.8868838548660278,
+      "learning_rate": 7.95708e-07,
+      "loss": 0.5589,
+      "step": 51100
+    },
+    {
+      "epoch": 0.5115,
+      "grad_norm": 48.44756317138672,
+      "learning_rate": 7.95508e-07,
+      "loss": 0.7304,
+      "step": 51150
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 36.71236038208008,
+      "learning_rate": 7.953079999999999e-07,
+      "loss": 0.5783,
+      "step": 51200
+    },
+    {
+      "epoch": 0.5125,
+      "grad_norm": 59.711605072021484,
+      "learning_rate": 7.95108e-07,
+      "loss": 0.5104,
+      "step": 51250
+    },
+    {
+      "epoch": 0.513,
+      "grad_norm": 21.43183135986328,
+      "learning_rate": 7.949079999999999e-07,
+      "loss": 0.7175,
+      "step": 51300
+    },
+    {
+      "epoch": 0.5135,
+      "grad_norm": 106.07775115966797,
+      "learning_rate": 7.947079999999999e-07,
+      "loss": 0.5952,
+      "step": 51350
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 25.90865135192871,
+      "learning_rate": 7.94508e-07,
+      "loss": 0.5459,
+      "step": 51400
+    },
+    {
+      "epoch": 0.5145,
+      "grad_norm": 3.6912739276885986,
+      "learning_rate": 7.94308e-07,
+      "loss": 0.5168,
+      "step": 51450
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 89.33435821533203,
+      "learning_rate": 7.94108e-07,
+      "loss": 0.5239,
+      "step": 51500
+    },
+    {
+      "epoch": 0.5155,
+      "grad_norm": 32.307071685791016,
+      "learning_rate": 7.939079999999999e-07,
+      "loss": 0.5169,
+      "step": 51550
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 16.09387969970703,
+      "learning_rate": 7.937079999999999e-07,
+      "loss": 0.4644,
+      "step": 51600
+    },
+    {
+      "epoch": 0.5165,
+      "grad_norm": 11.74305534362793,
+      "learning_rate": 7.935079999999999e-07,
+      "loss": 0.3657,
+      "step": 51650
+    },
+    {
+      "epoch": 0.517,
+      "grad_norm": 3.9428036212921143,
+      "learning_rate": 7.93308e-07,
+      "loss": 0.4922,
+      "step": 51700
+    },
+    {
+      "epoch": 0.5175,
+      "grad_norm": 15.612861633300781,
+      "learning_rate": 7.93108e-07,
+      "loss": 0.3998,
+      "step": 51750
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 96.3296890258789,
+      "learning_rate": 7.929079999999999e-07,
+      "loss": 0.5107,
+      "step": 51800
+    },
+    {
+      "epoch": 0.5185,
+      "grad_norm": 76.5985336303711,
+      "learning_rate": 7.92708e-07,
+      "loss": 0.5332,
+      "step": 51850
+    },
+    {
+      "epoch": 0.519,
+      "grad_norm": 17.58921241760254,
+      "learning_rate": 7.925079999999999e-07,
+      "loss": 0.4055,
+      "step": 51900
+    },
+    {
+      "epoch": 0.5195,
+      "grad_norm": 15.568597793579102,
+      "learning_rate": 7.923079999999999e-07,
+      "loss": 0.5198,
+      "step": 51950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8958696126937866,
+      "learning_rate": 7.92108e-07,
+      "loss": 0.5061,
+      "step": 52000
+    },
+    {
+      "epoch": 0.5205,
+      "grad_norm": 1.784369945526123,
+      "learning_rate": 7.919079999999999e-07,
+      "loss": 0.5103,
+      "step": 52050
+    },
+    {
+      "epoch": 0.521,
+      "grad_norm": 36.33938980102539,
+      "learning_rate": 7.917079999999999e-07,
+      "loss": 0.574,
+      "step": 52100
+    },
+    {
+      "epoch": 0.5215,
+      "grad_norm": 39.72173309326172,
+      "learning_rate": 7.91508e-07,
+      "loss": 0.5163,
+      "step": 52150
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 56.958919525146484,
+      "learning_rate": 7.91308e-07,
+      "loss": 0.5032,
+      "step": 52200
+    },
+    {
+      "epoch": 0.5225,
+      "grad_norm": 103.39403533935547,
+      "learning_rate": 7.91108e-07,
+      "loss": 0.63,
+      "step": 52250
+    },
+    {
+      "epoch": 0.523,
+      "grad_norm": 79.32686614990234,
+      "learning_rate": 7.90908e-07,
+      "loss": 0.5285,
+      "step": 52300
+    },
+    {
+      "epoch": 0.5235,
+      "grad_norm": 38.18978500366211,
+      "learning_rate": 7.907079999999999e-07,
+      "loss": 0.4923,
+      "step": 52350
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 17.309484481811523,
+      "learning_rate": 7.90508e-07,
+      "loss": 0.4923,
+      "step": 52400
+    },
+    {
+      "epoch": 0.5245,
+      "grad_norm": 92.0141372680664,
+      "learning_rate": 7.90308e-07,
+      "loss": 0.5748,
+      "step": 52450
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 38.47359848022461,
+      "learning_rate": 7.90108e-07,
+      "loss": 0.5513,
+      "step": 52500
+    },
+    {
+      "epoch": 0.5255,
+      "grad_norm": 69.26992797851562,
+      "learning_rate": 7.899080000000001e-07,
+      "loss": 0.6091,
+      "step": 52550
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 15.796204566955566,
+      "learning_rate": 7.897079999999999e-07,
+      "loss": 0.5999,
+      "step": 52600
+    },
+    {
+      "epoch": 0.5265,
+      "grad_norm": 103.76625061035156,
+      "learning_rate": 7.895079999999999e-07,
+      "loss": 0.4535,
+      "step": 52650
+    },
+    {
+      "epoch": 0.527,
+      "grad_norm": 6.661074161529541,
+      "learning_rate": 7.89308e-07,
+      "loss": 0.5645,
+      "step": 52700
+    },
+    {
+      "epoch": 0.5275,
+      "grad_norm": 92.19075012207031,
+      "learning_rate": 7.89108e-07,
+      "loss": 0.5311,
+      "step": 52750
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 51.94999313354492,
+      "learning_rate": 7.88908e-07,
+      "loss": 0.4162,
+      "step": 52800
+    },
+    {
+      "epoch": 0.5285,
+      "grad_norm": 2.7024729251861572,
+      "learning_rate": 7.88708e-07,
+      "loss": 0.4668,
+      "step": 52850
+    },
+    {
+      "epoch": 0.529,
+      "grad_norm": 72.6579360961914,
+      "learning_rate": 7.88508e-07,
+      "loss": 0.5353,
+      "step": 52900
+    },
+    {
+      "epoch": 0.5295,
+      "grad_norm": 47.76490783691406,
+      "learning_rate": 7.883079999999999e-07,
+      "loss": 0.3945,
+      "step": 52950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 64.03026580810547,
+      "learning_rate": 7.88108e-07,
+      "loss": 0.5546,
+      "step": 53000
+    },
+    {
+      "epoch": 0.5305,
+      "grad_norm": 37.60674285888672,
+      "learning_rate": 7.87908e-07,
+      "loss": 0.5378,
+      "step": 53050
+    },
+    {
+      "epoch": 0.531,
+      "grad_norm": 53.003170013427734,
+      "learning_rate": 7.877079999999999e-07,
+      "loss": 0.4513,
+      "step": 53100
+    },
+    {
+      "epoch": 0.5315,
+      "grad_norm": 5.38227653503418,
+      "learning_rate": 7.87508e-07,
+      "loss": 0.5933,
+      "step": 53150
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 48.471832275390625,
+      "learning_rate": 7.87308e-07,
+      "loss": 0.476,
+      "step": 53200
+    },
+    {
+      "epoch": 0.5325,
+      "grad_norm": 7.869255542755127,
+      "learning_rate": 7.87108e-07,
+      "loss": 0.4092,
+      "step": 53250
+    },
+    {
+      "epoch": 0.533,
+      "grad_norm": 18.027494430541992,
+      "learning_rate": 7.86908e-07,
+      "loss": 0.5511,
+      "step": 53300
+    },
+    {
+      "epoch": 0.5335,
+      "grad_norm": 2.7823119163513184,
+      "learning_rate": 7.867079999999999e-07,
+      "loss": 0.4462,
+      "step": 53350
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 84.66714477539062,
+      "learning_rate": 7.865079999999999e-07,
+      "loss": 0.5392,
+      "step": 53400
+    },
+    {
+      "epoch": 0.5345,
+      "grad_norm": 39.51984786987305,
+      "learning_rate": 7.86308e-07,
+      "loss": 0.5594,
+      "step": 53450
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 58.15394973754883,
+      "learning_rate": 7.86112e-07,
+      "loss": 0.49,
+      "step": 53500
+    },
+    {
+      "epoch": 0.5355,
+      "grad_norm": 90.69095611572266,
+      "learning_rate": 7.85912e-07,
+      "loss": 0.3737,
+      "step": 53550
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 6.2065629959106445,
+      "learning_rate": 7.85712e-07,
+      "loss": 0.6176,
+      "step": 53600
+    },
+    {
+      "epoch": 0.5365,
+      "grad_norm": 123.52501678466797,
+      "learning_rate": 7.855119999999999e-07,
+      "loss": 0.4808,
+      "step": 53650
+    },
+    {
+      "epoch": 0.537,
+      "grad_norm": 122.08316040039062,
+      "learning_rate": 7.853119999999999e-07,
+      "loss": 0.5169,
+      "step": 53700
+    },
+    {
+      "epoch": 0.5375,
+      "grad_norm": 1.263464331626892,
+      "learning_rate": 7.85112e-07,
+      "loss": 0.4527,
+      "step": 53750
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 0.7644511461257935,
+      "learning_rate": 7.84912e-07,
+      "loss": 0.6983,
+      "step": 53800
+    },
+    {
+      "epoch": 0.5385,
+      "grad_norm": 4.055157661437988,
+      "learning_rate": 7.847119999999999e-07,
+      "loss": 0.5582,
+      "step": 53850
+    },
+    {
+      "epoch": 0.539,
+      "grad_norm": 39.59327697753906,
+      "learning_rate": 7.84512e-07,
+      "loss": 0.4857,
+      "step": 53900
+    },
+    {
+      "epoch": 0.5395,
+      "grad_norm": 44.580570220947266,
+      "learning_rate": 7.84312e-07,
+      "loss": 0.5478,
+      "step": 53950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 45.88383865356445,
+      "learning_rate": 7.84112e-07,
+      "loss": 0.5136,
+      "step": 54000
+    },
+    {
+      "epoch": 0.5405,
+      "grad_norm": 18.86288833618164,
+      "learning_rate": 7.83912e-07,
+      "loss": 0.5359,
+      "step": 54050
+    },
+    {
+      "epoch": 0.541,
+      "grad_norm": 78.80966186523438,
+      "learning_rate": 7.837119999999999e-07,
+      "loss": 0.6137,
+      "step": 54100
+    },
+    {
+      "epoch": 0.5415,
+      "grad_norm": 17.388288497924805,
+      "learning_rate": 7.835119999999999e-07,
+      "loss": 0.5367,
+      "step": 54150
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 50.73162078857422,
+      "learning_rate": 7.83312e-07,
+      "loss": 0.5237,
+      "step": 54200
+    },
+    {
+      "epoch": 0.5425,
+      "grad_norm": 11.157291412353516,
+      "learning_rate": 7.83112e-07,
+      "loss": 0.4897,
+      "step": 54250
+    },
+    {
+      "epoch": 0.543,
+      "grad_norm": 75.02784729003906,
+      "learning_rate": 7.829120000000001e-07,
+      "loss": 0.5842,
+      "step": 54300
+    },
+    {
+      "epoch": 0.5435,
+      "grad_norm": 72.39640045166016,
+      "learning_rate": 7.82712e-07,
+      "loss": 0.5354,
+      "step": 54350
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.02550535462796688,
+      "learning_rate": 7.825119999999999e-07,
+      "loss": 0.5047,
+      "step": 54400
+    },
+    {
+      "epoch": 0.5445,
+      "grad_norm": 5.766872882843018,
+      "learning_rate": 7.82312e-07,
+      "loss": 0.5554,
+      "step": 54450
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 41.95703125,
+      "learning_rate": 7.82112e-07,
+      "loss": 0.5145,
+      "step": 54500
+    },
+    {
+      "epoch": 0.5455,
+      "grad_norm": 100.95028686523438,
+      "learning_rate": 7.81912e-07,
+      "loss": 0.5323,
+      "step": 54550
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 107.0252685546875,
+      "learning_rate": 7.81712e-07,
+      "loss": 0.6051,
+      "step": 54600
+    },
+    {
+      "epoch": 0.5465,
+      "grad_norm": 73.3103256225586,
+      "learning_rate": 7.81512e-07,
+      "loss": 0.6191,
+      "step": 54650
+    },
+    {
+      "epoch": 0.547,
+      "grad_norm": 99.1833724975586,
+      "learning_rate": 7.81312e-07,
+      "loss": 0.5797,
+      "step": 54700
+    },
+    {
+      "epoch": 0.5475,
+      "grad_norm": 64.93377685546875,
+      "learning_rate": 7.81112e-07,
+      "loss": 0.4962,
+      "step": 54750
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 16.584720611572266,
+      "learning_rate": 7.80912e-07,
+      "loss": 0.5693,
+      "step": 54800
+    },
+    {
+      "epoch": 0.5485,
+      "grad_norm": 91.99310302734375,
+      "learning_rate": 7.807119999999999e-07,
+      "loss": 0.6053,
+      "step": 54850
+    },
+    {
+      "epoch": 0.549,
+      "grad_norm": 91.74974822998047,
+      "learning_rate": 7.80512e-07,
+      "loss": 0.515,
+      "step": 54900
+    },
+    {
+      "epoch": 0.5495,
+      "grad_norm": 11.728752136230469,
+      "learning_rate": 7.80312e-07,
+      "loss": 0.499,
+      "step": 54950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.06698989868164,
+      "learning_rate": 7.80112e-07,
+      "loss": 0.4463,
+      "step": 55000
+    },
+    {
+      "epoch": 0.5505,
+      "grad_norm": 88.770751953125,
+      "learning_rate": 7.799120000000001e-07,
+      "loss": 0.5427,
+      "step": 55050
+    },
+    {
+      "epoch": 0.551,
+      "grad_norm": 27.026084899902344,
+      "learning_rate": 7.797119999999999e-07,
+      "loss": 0.523,
+      "step": 55100
+    },
+    {
+      "epoch": 0.5515,
+      "grad_norm": 1.4966083765029907,
+      "learning_rate": 7.795119999999999e-07,
+      "loss": 0.3774,
+      "step": 55150
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 48.237422943115234,
+      "learning_rate": 7.79312e-07,
+      "loss": 0.5175,
+      "step": 55200
+    },
+    {
+      "epoch": 0.5525,
+      "grad_norm": 55.568851470947266,
+      "learning_rate": 7.79112e-07,
+      "loss": 0.4215,
+      "step": 55250
+    },
+    {
+      "epoch": 0.553,
+      "grad_norm": 102.37129211425781,
+      "learning_rate": 7.78912e-07,
+      "loss": 0.64,
+      "step": 55300
+    },
+    {
+      "epoch": 0.5535,
+      "grad_norm": 3.732124090194702,
+      "learning_rate": 7.78712e-07,
+      "loss": 0.5539,
+      "step": 55350
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 49.1021842956543,
+      "learning_rate": 7.78512e-07,
+      "loss": 0.4958,
+      "step": 55400
+    },
+    {
+      "epoch": 0.5545,
+      "grad_norm": 85.12141418457031,
+      "learning_rate": 7.783119999999999e-07,
+      "loss": 0.4048,
+      "step": 55450
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 7.854513168334961,
+      "learning_rate": 7.78112e-07,
+      "loss": 0.4996,
+      "step": 55500
+    },
+    {
+      "epoch": 0.5555,
+      "grad_norm": 21.370290756225586,
+      "learning_rate": 7.77912e-07,
+      "loss": 0.6001,
+      "step": 55550
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 58.50725173950195,
+      "learning_rate": 7.777119999999999e-07,
+      "loss": 0.5398,
+      "step": 55600
+    },
+    {
+      "epoch": 0.5565,
+      "grad_norm": 26.769861221313477,
+      "learning_rate": 7.77512e-07,
+      "loss": 0.5714,
+      "step": 55650
+    },
+    {
+      "epoch": 0.557,
+      "grad_norm": 7.172065734863281,
+      "learning_rate": 7.77312e-07,
+      "loss": 0.5132,
+      "step": 55700
+    },
+    {
+      "epoch": 0.5575,
+      "grad_norm": 28.10909080505371,
+      "learning_rate": 7.77112e-07,
+      "loss": 0.5079,
+      "step": 55750
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 78.68544006347656,
+      "learning_rate": 7.76912e-07,
+      "loss": 0.575,
+      "step": 55800
+    },
+    {
+      "epoch": 0.5585,
+      "grad_norm": 105.1363754272461,
+      "learning_rate": 7.767119999999999e-07,
+      "loss": 0.6106,
+      "step": 55850
+    },
+    {
+      "epoch": 0.559,
+      "grad_norm": 67.24320983886719,
+      "learning_rate": 7.765159999999999e-07,
+      "loss": 0.5576,
+      "step": 55900
+    },
+    {
+      "epoch": 0.5595,
+      "grad_norm": 8.551283836364746,
+      "learning_rate": 7.76316e-07,
+      "loss": 0.6247,
+      "step": 55950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 33.97761535644531,
+      "learning_rate": 7.76116e-07,
+      "loss": 0.5779,
+      "step": 56000
+    },
+    {
+      "epoch": 0.5605,
+      "grad_norm": 106.0393295288086,
+      "learning_rate": 7.75916e-07,
+      "loss": 0.5783,
+      "step": 56050
+    },
+    {
+      "epoch": 0.561,
+      "grad_norm": 66.49019622802734,
+      "learning_rate": 7.75716e-07,
+      "loss": 0.5892,
+      "step": 56100
+    },
+    {
+      "epoch": 0.5615,
+      "grad_norm": 70.9111557006836,
+      "learning_rate": 7.755159999999999e-07,
+      "loss": 0.521,
+      "step": 56150
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 27.3460750579834,
+      "learning_rate": 7.753159999999999e-07,
+      "loss": 0.4994,
+      "step": 56200
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 26.073501586914062,
+      "learning_rate": 7.75116e-07,
+      "loss": 0.4284,
+      "step": 56250
+    },
+    {
+      "epoch": 0.563,
+      "grad_norm": 10.379983901977539,
+      "learning_rate": 7.74916e-07,
+      "loss": 0.4984,
+      "step": 56300
+    },
+    {
+      "epoch": 0.5635,
+      "grad_norm": 15.902315139770508,
+      "learning_rate": 7.747159999999999e-07,
+      "loss": 0.5109,
+      "step": 56350
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 13.556990623474121,
+      "learning_rate": 7.74516e-07,
+      "loss": 0.5017,
+      "step": 56400
+    },
+    {
+      "epoch": 0.5645,
+      "grad_norm": 32.288238525390625,
+      "learning_rate": 7.74316e-07,
+      "loss": 0.4915,
+      "step": 56450
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 82.15554809570312,
+      "learning_rate": 7.741159999999999e-07,
+      "loss": 0.5898,
+      "step": 56500
+    },
+    {
+      "epoch": 0.5655,
+      "grad_norm": 2.104949951171875,
+      "learning_rate": 7.73916e-07,
+      "loss": 0.4796,
+      "step": 56550
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 18.454978942871094,
+      "learning_rate": 7.737159999999999e-07,
+      "loss": 0.646,
+      "step": 56600
+    },
+    {
+      "epoch": 0.5665,
+      "grad_norm": 1.2304694652557373,
+      "learning_rate": 7.735159999999999e-07,
+      "loss": 0.5653,
+      "step": 56650
+    },
+    {
+      "epoch": 0.567,
+      "grad_norm": 87.61625671386719,
+      "learning_rate": 7.73316e-07,
+      "loss": 0.4816,
+      "step": 56700
+    },
+    {
+      "epoch": 0.5675,
+      "grad_norm": 39.44472885131836,
+      "learning_rate": 7.73116e-07,
+      "loss": 0.4758,
+      "step": 56750
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 60.77321243286133,
+      "learning_rate": 7.729160000000001e-07,
+      "loss": 0.5649,
+      "step": 56800
+    },
+    {
+      "epoch": 0.5685,
+      "grad_norm": 62.973541259765625,
+      "learning_rate": 7.727159999999999e-07,
+      "loss": 0.5331,
+      "step": 56850
+    },
+    {
+      "epoch": 0.569,
+      "grad_norm": 3.172307252883911,
+      "learning_rate": 7.725159999999999e-07,
+      "loss": 0.529,
+      "step": 56900
+    },
+    {
+      "epoch": 0.5695,
+      "grad_norm": 82.92523956298828,
+      "learning_rate": 7.72316e-07,
+      "loss": 0.4194,
+      "step": 56950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 30.7186222076416,
+      "learning_rate": 7.72116e-07,
+      "loss": 0.4916,
+      "step": 57000
+    },
+    {
+      "epoch": 0.5705,
+      "grad_norm": 21.11612319946289,
+      "learning_rate": 7.71916e-07,
+      "loss": 0.5155,
+      "step": 57050
+    },
+    {
+      "epoch": 0.571,
+      "grad_norm": 109.33685302734375,
+      "learning_rate": 7.71716e-07,
+      "loss": 0.6148,
+      "step": 57100
+    },
+    {
+      "epoch": 0.5715,
+      "grad_norm": 4.589968204498291,
+      "learning_rate": 7.71516e-07,
+      "loss": 0.4523,
+      "step": 57150
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 50.23842239379883,
+      "learning_rate": 7.713159999999999e-07,
+      "loss": 0.4689,
+      "step": 57200
+    },
+    {
+      "epoch": 0.5725,
+      "grad_norm": 5.932251453399658,
+      "learning_rate": 7.71116e-07,
+      "loss": 0.4741,
+      "step": 57250
+    },
+    {
+      "epoch": 0.573,
+      "grad_norm": 83.40533447265625,
+      "learning_rate": 7.70916e-07,
+      "loss": 0.522,
+      "step": 57300
+    },
+    {
+      "epoch": 0.5735,
+      "grad_norm": 53.40946960449219,
+      "learning_rate": 7.707159999999999e-07,
+      "loss": 0.3644,
+      "step": 57350
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 107.14583587646484,
+      "learning_rate": 7.70516e-07,
+      "loss": 0.4899,
+      "step": 57400
+    },
+    {
+      "epoch": 0.5745,
+      "grad_norm": 81.80699920654297,
+      "learning_rate": 7.70316e-07,
+      "loss": 0.4938,
+      "step": 57450
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 66.33548736572266,
+      "learning_rate": 7.70116e-07,
+      "loss": 0.5103,
+      "step": 57500
+    },
+    {
+      "epoch": 0.5755,
+      "grad_norm": 55.853206634521484,
+      "learning_rate": 7.69916e-07,
+      "loss": 0.5101,
+      "step": 57550
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.4973876476287842,
+      "learning_rate": 7.697159999999999e-07,
+      "loss": 0.5066,
+      "step": 57600
+    },
+    {
+      "epoch": 0.5765,
+      "grad_norm": 8.70065689086914,
+      "learning_rate": 7.695159999999999e-07,
+      "loss": 0.523,
+      "step": 57650
+    },
+    {
+      "epoch": 0.577,
+      "grad_norm": 68.1734390258789,
+      "learning_rate": 7.69316e-07,
+      "loss": 0.5269,
+      "step": 57700
+    },
+    {
+      "epoch": 0.5775,
+      "grad_norm": 47.97027587890625,
+      "learning_rate": 7.69116e-07,
+      "loss": 0.4298,
+      "step": 57750
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 0.9205309152603149,
+      "learning_rate": 7.68916e-07,
+      "loss": 0.4194,
+      "step": 57800
+    },
+    {
+      "epoch": 0.5785,
+      "grad_norm": 73.67825317382812,
+      "learning_rate": 7.68716e-07,
+      "loss": 0.4472,
+      "step": 57850
+    },
+    {
+      "epoch": 0.579,
+      "grad_norm": 41.70149612426758,
+      "learning_rate": 7.685159999999999e-07,
+      "loss": 0.4958,
+      "step": 57900
+    },
+    {
+      "epoch": 0.5795,
+      "grad_norm": 37.12933349609375,
+      "learning_rate": 7.683159999999999e-07,
+      "loss": 0.3652,
+      "step": 57950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 50.90730667114258,
+      "learning_rate": 7.68116e-07,
+      "loss": 0.4357,
+      "step": 58000
+    },
+    {
+      "epoch": 0.5805,
+      "grad_norm": 84.36893463134766,
+      "learning_rate": 7.67916e-07,
+      "loss": 0.643,
+      "step": 58050
+    },
+    {
+      "epoch": 0.581,
+      "grad_norm": 2.3979945182800293,
+      "learning_rate": 7.677159999999999e-07,
+      "loss": 0.4654,
+      "step": 58100
+    },
+    {
+      "epoch": 0.5815,
+      "grad_norm": 82.98129272460938,
+      "learning_rate": 7.67516e-07,
+      "loss": 0.5124,
+      "step": 58150
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 68.0512924194336,
+      "learning_rate": 7.67316e-07,
+      "loss": 0.5489,
+      "step": 58200
+    },
+    {
+      "epoch": 0.5825,
+      "grad_norm": 7.42828893661499,
+      "learning_rate": 7.671159999999999e-07,
+      "loss": 0.6616,
+      "step": 58250
+    },
+    {
+      "epoch": 0.583,
+      "grad_norm": 16.994314193725586,
+      "learning_rate": 7.66916e-07,
+      "loss": 0.4413,
+      "step": 58300
+    },
+    {
+      "epoch": 0.5835,
+      "grad_norm": 33.765499114990234,
+      "learning_rate": 7.667159999999999e-07,
+      "loss": 0.5723,
+      "step": 58350
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 114.72591400146484,
+      "learning_rate": 7.665159999999999e-07,
+      "loss": 0.4305,
+      "step": 58400
+    },
+    {
+      "epoch": 0.5845,
+      "grad_norm": 60.763423919677734,
+      "learning_rate": 7.66316e-07,
+      "loss": 0.4434,
+      "step": 58450
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 70.40579223632812,
+      "learning_rate": 7.66116e-07,
+      "loss": 0.5024,
+      "step": 58500
+    },
+    {
+      "epoch": 0.5855,
+      "grad_norm": 63.25678634643555,
+      "learning_rate": 7.65916e-07,
+      "loss": 0.4624,
+      "step": 58550
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 75.37684631347656,
+      "learning_rate": 7.657159999999999e-07,
+      "loss": 0.6335,
+      "step": 58600
+    },
+    {
+      "epoch": 0.5865,
+      "grad_norm": 82.71162414550781,
+      "learning_rate": 7.655159999999999e-07,
+      "loss": 0.57,
+      "step": 58650
+    },
+    {
+      "epoch": 0.587,
+      "grad_norm": 6.6550445556640625,
+      "learning_rate": 7.653159999999999e-07,
+      "loss": 0.6168,
+      "step": 58700
+    },
+    {
+      "epoch": 0.5875,
+      "grad_norm": 21.841920852661133,
+      "learning_rate": 7.65116e-07,
+      "loss": 0.4213,
+      "step": 58750
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 67.11859893798828,
+      "learning_rate": 7.64916e-07,
+      "loss": 0.5094,
+      "step": 58800
+    },
+    {
+      "epoch": 0.5885,
+      "grad_norm": 6.230730056762695,
+      "learning_rate": 7.647159999999999e-07,
+      "loss": 0.5381,
+      "step": 58850
+    },
+    {
+      "epoch": 0.589,
+      "grad_norm": 90.90957641601562,
+      "learning_rate": 7.64516e-07,
+      "loss": 0.5987,
+      "step": 58900
+    },
+    {
+      "epoch": 0.5895,
+      "grad_norm": 67.64148712158203,
+      "learning_rate": 7.643159999999999e-07,
+      "loss": 0.5335,
+      "step": 58950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 82.5676040649414,
+      "learning_rate": 7.641159999999999e-07,
+      "loss": 0.524,
+      "step": 59000
+    },
+    {
+      "epoch": 0.5905,
+      "grad_norm": 2.6406068801879883,
+      "learning_rate": 7.63916e-07,
+      "loss": 0.4817,
+      "step": 59050
+    },
+    {
+      "epoch": 0.591,
+      "grad_norm": 44.6208381652832,
+      "learning_rate": 7.637159999999999e-07,
+      "loss": 0.5026,
+      "step": 59100
+    },
+    {
+      "epoch": 0.5915,
+      "grad_norm": 90.23446655273438,
+      "learning_rate": 7.635159999999999e-07,
+      "loss": 0.5775,
+      "step": 59150
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 1.308499813079834,
+      "learning_rate": 7.63316e-07,
+      "loss": 0.5328,
+      "step": 59200
+    },
+    {
+      "epoch": 0.5925,
+      "grad_norm": 97.81929016113281,
+      "learning_rate": 7.63116e-07,
+      "loss": 0.5452,
+      "step": 59250
+    },
+    {
+      "epoch": 0.593,
+      "grad_norm": 21.65294647216797,
+      "learning_rate": 7.629160000000001e-07,
+      "loss": 0.5259,
+      "step": 59300
+    },
+    {
+      "epoch": 0.5935,
+      "grad_norm": 15.141357421875,
+      "learning_rate": 7.627159999999999e-07,
+      "loss": 0.3815,
+      "step": 59350
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 62.881248474121094,
+      "learning_rate": 7.625159999999999e-07,
+      "loss": 0.5372,
+      "step": 59400
+    },
+    {
+      "epoch": 0.5945,
+      "grad_norm": 2.6095540523529053,
+      "learning_rate": 7.62316e-07,
+      "loss": 0.4359,
+      "step": 59450
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 5.5010085105896,
+      "learning_rate": 7.62116e-07,
+      "loss": 0.4372,
+      "step": 59500
+    },
+    {
+      "epoch": 0.5955,
+      "grad_norm": 35.81318664550781,
+      "learning_rate": 7.61916e-07,
+      "loss": 0.442,
+      "step": 59550
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 61.836212158203125,
+      "learning_rate": 7.61716e-07,
+      "loss": 0.6257,
+      "step": 59600
+    },
+    {
+      "epoch": 0.5965,
+      "grad_norm": 89.80731201171875,
+      "learning_rate": 7.61516e-07,
+      "loss": 0.5622,
+      "step": 59650
+    },
+    {
+      "epoch": 0.597,
+      "grad_norm": 59.49220657348633,
+      "learning_rate": 7.613159999999999e-07,
+      "loss": 0.4933,
+      "step": 59700
+    },
+    {
+      "epoch": 0.5975,
+      "grad_norm": 64.80416107177734,
+      "learning_rate": 7.61116e-07,
+      "loss": 0.6054,
+      "step": 59750
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 42.769248962402344,
+      "learning_rate": 7.60916e-07,
+      "loss": 0.4669,
+      "step": 59800
+    },
+    {
+      "epoch": 0.5985,
+      "grad_norm": 4.290161609649658,
+      "learning_rate": 7.607159999999999e-07,
+      "loss": 0.5286,
+      "step": 59850
+    },
+    {
+      "epoch": 0.599,
+      "grad_norm": 50.70981216430664,
+      "learning_rate": 7.60516e-07,
+      "loss": 0.5784,
+      "step": 59900
+    },
+    {
+      "epoch": 0.5995,
+      "grad_norm": 64.66263580322266,
+      "learning_rate": 7.60316e-07,
+      "loss": 0.558,
+      "step": 59950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 56.63187026977539,
+      "learning_rate": 7.60116e-07,
+      "loss": 0.59,
+      "step": 60000
+    },
+    {
+      "epoch": 0.6005,
+      "grad_norm": 53.35725402832031,
+      "learning_rate": 7.59916e-07,
+      "loss": 0.6365,
+      "step": 60050
+    },
+    {
+      "epoch": 0.601,
+      "grad_norm": 30.334291458129883,
+      "learning_rate": 7.597159999999999e-07,
+      "loss": 0.5164,
+      "step": 60100
+    },
+    {
+      "epoch": 0.6015,
+      "grad_norm": 9.290550231933594,
+      "learning_rate": 7.595159999999999e-07,
+      "loss": 0.515,
+      "step": 60150
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 64.14176940917969,
+      "learning_rate": 7.59316e-07,
+      "loss": 0.614,
+      "step": 60200
+    },
+    {
+      "epoch": 0.6025,
+      "grad_norm": 75.271484375,
+      "learning_rate": 7.59116e-07,
+      "loss": 0.6181,
+      "step": 60250
+    },
+    {
+      "epoch": 0.603,
+      "grad_norm": 4.262742042541504,
+      "learning_rate": 7.58916e-07,
+      "loss": 0.5013,
+      "step": 60300
+    },
+    {
+      "epoch": 0.6035,
+      "grad_norm": 3.081693649291992,
+      "learning_rate": 7.58716e-07,
+      "loss": 0.5908,
+      "step": 60350
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 39.718345642089844,
+      "learning_rate": 7.585159999999999e-07,
+      "loss": 0.4868,
+      "step": 60400
+    },
+    {
+      "epoch": 0.6045,
+      "grad_norm": 84.18822479248047,
+      "learning_rate": 7.583159999999999e-07,
+      "loss": 0.5012,
+      "step": 60450
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 90.46514892578125,
+      "learning_rate": 7.58116e-07,
+      "loss": 0.5695,
+      "step": 60500
+    },
+    {
+      "epoch": 0.6055,
+      "grad_norm": 56.23940658569336,
+      "learning_rate": 7.57916e-07,
+      "loss": 0.5034,
+      "step": 60550
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 27.244321823120117,
+      "learning_rate": 7.577159999999999e-07,
+      "loss": 0.5065,
+      "step": 60600
+    },
+    {
+      "epoch": 0.6065,
+      "grad_norm": 6.311286449432373,
+      "learning_rate": 7.57516e-07,
+      "loss": 0.5169,
+      "step": 60650
+    },
+    {
+      "epoch": 0.607,
+      "grad_norm": 1.7884770631790161,
+      "learning_rate": 7.57316e-07,
+      "loss": 0.4025,
+      "step": 60700
+    },
+    {
+      "epoch": 0.6075,
+      "grad_norm": 76.4052734375,
+      "learning_rate": 7.571159999999999e-07,
+      "loss": 0.5358,
+      "step": 60750
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 57.62378692626953,
+      "learning_rate": 7.56916e-07,
+      "loss": 0.5542,
+      "step": 60800
+    },
+    {
+      "epoch": 0.6085,
+      "grad_norm": 1.3535525798797607,
+      "learning_rate": 7.56716e-07,
+      "loss": 0.6046,
+      "step": 60850
+    },
+    {
+      "epoch": 0.609,
+      "grad_norm": 57.33085632324219,
+      "learning_rate": 7.565159999999999e-07,
+      "loss": 0.5285,
+      "step": 60900
+    },
+    {
+      "epoch": 0.6095,
+      "grad_norm": 10.677825927734375,
+      "learning_rate": 7.56316e-07,
+      "loss": 0.4391,
+      "step": 60950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.921430587768555,
+      "learning_rate": 7.56116e-07,
+      "loss": 0.4501,
+      "step": 61000
+    },
+    {
+      "epoch": 0.6105,
+      "grad_norm": 3.2991600036621094,
+      "learning_rate": 7.55916e-07,
+      "loss": 0.5235,
+      "step": 61050
+    },
+    {
+      "epoch": 0.611,
+      "grad_norm": 50.49160385131836,
+      "learning_rate": 7.55716e-07,
+      "loss": 0.4383,
+      "step": 61100
+    },
+    {
+      "epoch": 0.6115,
+      "grad_norm": 93.91867065429688,
+      "learning_rate": 7.555159999999999e-07,
+      "loss": 0.558,
+      "step": 61150
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 15.711594581604004,
+      "learning_rate": 7.553159999999999e-07,
+      "loss": 0.5294,
+      "step": 61200
+    },
+    {
+      "epoch": 0.6125,
+      "grad_norm": 9.569193840026855,
+      "learning_rate": 7.5512e-07,
+      "loss": 0.4918,
+      "step": 61250
+    },
+    {
+      "epoch": 0.613,
+      "grad_norm": 36.533626556396484,
+      "learning_rate": 7.5492e-07,
+      "loss": 0.4458,
+      "step": 61300
+    },
+    {
+      "epoch": 0.6135,
+      "grad_norm": 47.60615158081055,
+      "learning_rate": 7.547199999999999e-07,
+      "loss": 0.494,
+      "step": 61350
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 47.9034538269043,
+      "learning_rate": 7.5452e-07,
+      "loss": 0.5418,
+      "step": 61400
+    },
+    {
+      "epoch": 0.6145,
+      "grad_norm": 55.0429573059082,
+      "learning_rate": 7.543199999999999e-07,
+      "loss": 0.6425,
+      "step": 61450
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 29.267080307006836,
+      "learning_rate": 7.541199999999999e-07,
+      "loss": 0.467,
+      "step": 61500
+    },
+    {
+      "epoch": 0.6155,
+      "grad_norm": 47.8558235168457,
+      "learning_rate": 7.5392e-07,
+      "loss": 0.4311,
+      "step": 61550
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 63.829071044921875,
+      "learning_rate": 7.537199999999999e-07,
+      "loss": 0.564,
+      "step": 61600
+    },
+    {
+      "epoch": 0.6165,
+      "grad_norm": 2.9663500785827637,
+      "learning_rate": 7.535199999999999e-07,
+      "loss": 0.5972,
+      "step": 61650
+    },
+    {
+      "epoch": 0.617,
+      "grad_norm": 1.3793628215789795,
+      "learning_rate": 7.5332e-07,
+      "loss": 0.4365,
+      "step": 61700
+    },
+    {
+      "epoch": 0.6175,
+      "grad_norm": 37.88240432739258,
+      "learning_rate": 7.5312e-07,
+      "loss": 0.596,
+      "step": 61750
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 71.33352661132812,
+      "learning_rate": 7.5292e-07,
+      "loss": 0.4351,
+      "step": 61800
+    },
+    {
+      "epoch": 0.6185,
+      "grad_norm": 59.38632583618164,
+      "learning_rate": 7.527199999999999e-07,
+      "loss": 0.4588,
+      "step": 61850
+    },
+    {
+      "epoch": 0.619,
+      "grad_norm": 59.172969818115234,
+      "learning_rate": 7.525199999999999e-07,
+      "loss": 0.5131,
+      "step": 61900
+    },
+    {
+      "epoch": 0.6195,
+      "grad_norm": 11.83280086517334,
+      "learning_rate": 7.5232e-07,
+      "loss": 0.5119,
+      "step": 61950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 64.36370849609375,
+      "learning_rate": 7.5212e-07,
+      "loss": 0.4334,
+      "step": 62000
+    },
+    {
+      "epoch": 0.6205,
+      "grad_norm": 63.02851867675781,
+      "learning_rate": 7.5192e-07,
+      "loss": 0.4878,
+      "step": 62050
+    },
+    {
+      "epoch": 0.621,
+      "grad_norm": 1.058779001235962,
+      "learning_rate": 7.517200000000001e-07,
+      "loss": 0.5113,
+      "step": 62100
+    },
+    {
+      "epoch": 0.6215,
+      "grad_norm": 2.1751315593719482,
+      "learning_rate": 7.515199999999999e-07,
+      "loss": 0.4869,
+      "step": 62150
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 100.90425109863281,
+      "learning_rate": 7.513199999999999e-07,
+      "loss": 0.5549,
+      "step": 62200
+    },
+    {
+      "epoch": 0.6225,
+      "grad_norm": 116.42280578613281,
+      "learning_rate": 7.5112e-07,
+      "loss": 0.4273,
+      "step": 62250
+    },
+    {
+      "epoch": 0.623,
+      "grad_norm": 81.34993743896484,
+      "learning_rate": 7.5092e-07,
+      "loss": 0.5264,
+      "step": 62300
+    },
+    {
+      "epoch": 0.6235,
+      "grad_norm": 27.30585289001465,
+      "learning_rate": 7.5072e-07,
+      "loss": 0.5647,
+      "step": 62350
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 8.460829734802246,
+      "learning_rate": 7.5052e-07,
+      "loss": 0.5671,
+      "step": 62400
+    },
+    {
+      "epoch": 0.6245,
+      "grad_norm": 23.786176681518555,
+      "learning_rate": 7.5032e-07,
+      "loss": 0.4409,
+      "step": 62450
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 100.85468292236328,
+      "learning_rate": 7.501199999999999e-07,
+      "loss": 0.5269,
+      "step": 62500
+    },
+    {
+      "epoch": 0.6255,
+      "grad_norm": 55.07974624633789,
+      "learning_rate": 7.4992e-07,
+      "loss": 0.3994,
+      "step": 62550
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 50.99776840209961,
+      "learning_rate": 7.4972e-07,
+      "loss": 0.5842,
+      "step": 62600
+    },
+    {
+      "epoch": 0.6265,
+      "grad_norm": 6.158398151397705,
+      "learning_rate": 7.495199999999999e-07,
+      "loss": 0.4597,
+      "step": 62650
+    },
+    {
+      "epoch": 0.627,
+      "grad_norm": 47.69706344604492,
+      "learning_rate": 7.4932e-07,
+      "loss": 0.5979,
+      "step": 62700
+    },
+    {
+      "epoch": 0.6275,
+      "grad_norm": 42.70079040527344,
+      "learning_rate": 7.4912e-07,
+      "loss": 0.4116,
+      "step": 62750
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 109.9066162109375,
+      "learning_rate": 7.4892e-07,
+      "loss": 0.4997,
+      "step": 62800
+    },
+    {
+      "epoch": 0.6285,
+      "grad_norm": 51.17443084716797,
+      "learning_rate": 7.4872e-07,
+      "loss": 0.5929,
+      "step": 62850
+    },
+    {
+      "epoch": 0.629,
+      "grad_norm": 24.46744155883789,
+      "learning_rate": 7.485199999999999e-07,
+      "loss": 0.5733,
+      "step": 62900
+    },
+    {
+      "epoch": 0.6295,
+      "grad_norm": 16.23428726196289,
+      "learning_rate": 7.483199999999999e-07,
+      "loss": 0.3725,
+      "step": 62950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.631766319274902,
+      "learning_rate": 7.4812e-07,
+      "loss": 0.5679,
+      "step": 63000
+    },
+    {
+      "epoch": 0.6305,
+      "grad_norm": 127.42314147949219,
+      "learning_rate": 7.4792e-07,
+      "loss": 0.4697,
+      "step": 63050
+    },
+    {
+      "epoch": 0.631,
+      "grad_norm": 85.11431121826172,
+      "learning_rate": 7.4772e-07,
+      "loss": 0.4926,
+      "step": 63100
+    },
+    {
+      "epoch": 0.6315,
+      "grad_norm": 18.65448570251465,
+      "learning_rate": 7.4752e-07,
+      "loss": 0.3696,
+      "step": 63150
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 52.924503326416016,
+      "learning_rate": 7.473199999999999e-07,
+      "loss": 0.6384,
+      "step": 63200
+    },
+    {
+      "epoch": 0.6325,
+      "grad_norm": 81.14250946044922,
+      "learning_rate": 7.471199999999999e-07,
+      "loss": 0.4821,
+      "step": 63250
+    },
+    {
+      "epoch": 0.633,
+      "grad_norm": 36.8980827331543,
+      "learning_rate": 7.4692e-07,
+      "loss": 0.456,
+      "step": 63300
+    },
+    {
+      "epoch": 0.6335,
+      "grad_norm": 39.605323791503906,
+      "learning_rate": 7.4672e-07,
+      "loss": 0.4922,
+      "step": 63350
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 31.341976165771484,
+      "learning_rate": 7.465199999999999e-07,
+      "loss": 0.466,
+      "step": 63400
+    },
+    {
+      "epoch": 0.6345,
+      "grad_norm": 87.95755004882812,
+      "learning_rate": 7.4632e-07,
+      "loss": 0.3872,
+      "step": 63450
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 59.00345993041992,
+      "learning_rate": 7.4612e-07,
+      "loss": 0.5011,
+      "step": 63500
+    },
+    {
+      "epoch": 0.6355,
+      "grad_norm": 31.053945541381836,
+      "learning_rate": 7.459199999999999e-07,
+      "loss": 0.6692,
+      "step": 63550
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 47.0079345703125,
+      "learning_rate": 7.45724e-07,
+      "loss": 0.6124,
+      "step": 63600
+    },
+    {
+      "epoch": 0.6365,
+      "grad_norm": 49.89374542236328,
+      "learning_rate": 7.455239999999999e-07,
+      "loss": 0.4711,
+      "step": 63650
+    },
+    {
+      "epoch": 0.637,
+      "grad_norm": 93.30340576171875,
+      "learning_rate": 7.453239999999999e-07,
+      "loss": 0.5194,
+      "step": 63700
+    },
+    {
+      "epoch": 0.6375,
+      "grad_norm": 95.00665283203125,
+      "learning_rate": 7.45124e-07,
+      "loss": 0.5065,
+      "step": 63750
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 50.189510345458984,
+      "learning_rate": 7.44924e-07,
+      "loss": 0.3856,
+      "step": 63800
+    },
+    {
+      "epoch": 0.6385,
+      "grad_norm": 37.73897933959961,
+      "learning_rate": 7.447240000000001e-07,
+      "loss": 0.4225,
+      "step": 63850
+    },
+    {
+      "epoch": 0.639,
+      "grad_norm": 28.45961570739746,
+      "learning_rate": 7.44524e-07,
+      "loss": 0.5747,
+      "step": 63900
+    },
+    {
+      "epoch": 0.6395,
+      "grad_norm": 44.61785888671875,
+      "learning_rate": 7.443239999999999e-07,
+      "loss": 0.6641,
+      "step": 63950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 48.99622344970703,
+      "learning_rate": 7.44124e-07,
+      "loss": 0.6024,
+      "step": 64000
+    },
+    {
+      "epoch": 0.6405,
+      "grad_norm": 22.556676864624023,
+      "learning_rate": 7.43924e-07,
+      "loss": 0.5411,
+      "step": 64050
+    },
+    {
+      "epoch": 0.641,
+      "grad_norm": 72.16722106933594,
+      "learning_rate": 7.43724e-07,
+      "loss": 0.5184,
+      "step": 64100
+    },
+    {
+      "epoch": 0.6415,
+      "grad_norm": 32.8302116394043,
+      "learning_rate": 7.43524e-07,
+      "loss": 0.4113,
+      "step": 64150
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 47.202213287353516,
+      "learning_rate": 7.43324e-07,
+      "loss": 0.6019,
+      "step": 64200
+    },
+    {
+      "epoch": 0.6425,
+      "grad_norm": 58.76848602294922,
+      "learning_rate": 7.43124e-07,
+      "loss": 0.5887,
+      "step": 64250
+    },
+    {
+      "epoch": 0.643,
+      "grad_norm": 54.78934097290039,
+      "learning_rate": 7.42924e-07,
+      "loss": 0.5743,
+      "step": 64300
+    },
+    {
+      "epoch": 0.6435,
+      "grad_norm": 16.40074348449707,
+      "learning_rate": 7.42724e-07,
+      "loss": 0.531,
+      "step": 64350
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 30.374088287353516,
+      "learning_rate": 7.425239999999999e-07,
+      "loss": 0.3552,
+      "step": 64400
+    },
+    {
+      "epoch": 0.6445,
+      "grad_norm": 14.19670295715332,
+      "learning_rate": 7.42324e-07,
+      "loss": 0.6266,
+      "step": 64450
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 86.16085052490234,
+      "learning_rate": 7.42124e-07,
+      "loss": 0.5935,
+      "step": 64500
+    },
+    {
+      "epoch": 0.6455,
+      "grad_norm": 84.91423797607422,
+      "learning_rate": 7.41924e-07,
+      "loss": 0.4733,
+      "step": 64550
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 72.64830780029297,
+      "learning_rate": 7.417240000000001e-07,
+      "loss": 0.4532,
+      "step": 64600
+    },
+    {
+      "epoch": 0.6465,
+      "grad_norm": 6.822329998016357,
+      "learning_rate": 7.415239999999999e-07,
+      "loss": 0.4741,
+      "step": 64650
+    },
+    {
+      "epoch": 0.647,
+      "grad_norm": 36.594993591308594,
+      "learning_rate": 7.413239999999999e-07,
+      "loss": 0.531,
+      "step": 64700
+    },
+    {
+      "epoch": 0.6475,
+      "grad_norm": 75.4520263671875,
+      "learning_rate": 7.41124e-07,
+      "loss": 0.4991,
+      "step": 64750
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 44.896141052246094,
+      "learning_rate": 7.40924e-07,
+      "loss": 0.4959,
+      "step": 64800
+    },
+    {
+      "epoch": 0.6485,
+      "grad_norm": 47.80727767944336,
+      "learning_rate": 7.40724e-07,
+      "loss": 0.5498,
+      "step": 64850
+    },
+    {
+      "epoch": 0.649,
+      "grad_norm": 2.48860502243042,
+      "learning_rate": 7.40524e-07,
+      "loss": 0.5012,
+      "step": 64900
+    },
+    {
+      "epoch": 0.6495,
+      "grad_norm": 16.638439178466797,
+      "learning_rate": 7.40324e-07,
+      "loss": 0.6177,
+      "step": 64950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 44.449302673339844,
+      "learning_rate": 7.401239999999999e-07,
+      "loss": 0.4293,
+      "step": 65000
+    },
+    {
+      "epoch": 0.6505,
+      "grad_norm": 161.3533935546875,
+      "learning_rate": 7.39924e-07,
+      "loss": 0.4377,
+      "step": 65050
+    },
+    {
+      "epoch": 0.651,
+      "grad_norm": 12.957307815551758,
+      "learning_rate": 7.39724e-07,
+      "loss": 0.5337,
+      "step": 65100
+    },
+    {
+      "epoch": 0.6515,
+      "grad_norm": 13.819578170776367,
+      "learning_rate": 7.395239999999999e-07,
+      "loss": 0.4665,
+      "step": 65150
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 108.1066665649414,
+      "learning_rate": 7.39324e-07,
+      "loss": 0.61,
+      "step": 65200
+    },
+    {
+      "epoch": 0.6525,
+      "grad_norm": 18.782867431640625,
+      "learning_rate": 7.39124e-07,
+      "loss": 0.5585,
+      "step": 65250
+    },
+    {
+      "epoch": 0.653,
+      "grad_norm": 14.486738204956055,
+      "learning_rate": 7.38924e-07,
+      "loss": 0.5132,
+      "step": 65300
+    },
+    {
+      "epoch": 0.6535,
+      "grad_norm": 28.351959228515625,
+      "learning_rate": 7.38724e-07,
+      "loss": 0.4924,
+      "step": 65350
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 49.564491271972656,
+      "learning_rate": 7.385239999999999e-07,
+      "loss": 0.485,
+      "step": 65400
+    },
+    {
+      "epoch": 0.6545,
+      "grad_norm": 55.472877502441406,
+      "learning_rate": 7.383239999999999e-07,
+      "loss": 0.5007,
+      "step": 65450
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 37.1424446105957,
+      "learning_rate": 7.38124e-07,
+      "loss": 0.518,
+      "step": 65500
+    },
+    {
+      "epoch": 0.6555,
+      "grad_norm": 30.962854385375977,
+      "learning_rate": 7.37924e-07,
+      "loss": 0.5072,
+      "step": 65550
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 63.9365348815918,
+      "learning_rate": 7.37724e-07,
+      "loss": 0.5801,
+      "step": 65600
+    },
+    {
+      "epoch": 0.6565,
+      "grad_norm": 6.695558547973633,
+      "learning_rate": 7.37524e-07,
+      "loss": 0.6023,
+      "step": 65650
+    },
+    {
+      "epoch": 0.657,
+      "grad_norm": 26.7724666595459,
+      "learning_rate": 7.373239999999999e-07,
+      "loss": 0.5249,
+      "step": 65700
+    },
+    {
+      "epoch": 0.6575,
+      "grad_norm": 81.64033508300781,
+      "learning_rate": 7.371239999999999e-07,
+      "loss": 0.5675,
+      "step": 65750
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 32.74658966064453,
+      "learning_rate": 7.36924e-07,
+      "loss": 0.5071,
+      "step": 65800
+    },
+    {
+      "epoch": 0.6585,
+      "grad_norm": 58.301597595214844,
+      "learning_rate": 7.36724e-07,
+      "loss": 0.5389,
+      "step": 65850
+    },
+    {
+      "epoch": 0.659,
+      "grad_norm": 53.7860221862793,
+      "learning_rate": 7.365279999999999e-07,
+      "loss": 0.4365,
+      "step": 65900
+    },
+    {
+      "epoch": 0.6595,
+      "grad_norm": 20.44268798828125,
+      "learning_rate": 7.36328e-07,
+      "loss": 0.4764,
+      "step": 65950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.711496353149414,
+      "learning_rate": 7.36128e-07,
+      "loss": 0.4768,
+      "step": 66000
+    },
+    {
+      "epoch": 0.6605,
+      "grad_norm": 61.34829330444336,
+      "learning_rate": 7.359279999999999e-07,
+      "loss": 0.3809,
+      "step": 66050
+    },
+    {
+      "epoch": 0.661,
+      "grad_norm": 95.46158599853516,
+      "learning_rate": 7.35732e-07,
+      "loss": 0.4626,
+      "step": 66100
+    },
+    {
+      "epoch": 0.6615,
+      "grad_norm": 50.04861068725586,
+      "learning_rate": 7.355319999999999e-07,
+      "loss": 0.6196,
+      "step": 66150
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 111.31083679199219,
+      "learning_rate": 7.353319999999999e-07,
+      "loss": 0.4418,
+      "step": 66200
+    },
+    {
+      "epoch": 0.6625,
+      "grad_norm": 0.3906168043613434,
+      "learning_rate": 7.35132e-07,
+      "loss": 0.3852,
+      "step": 66250
+    },
+    {
+      "epoch": 0.663,
+      "grad_norm": 45.034507751464844,
+      "learning_rate": 7.34932e-07,
+      "loss": 0.4873,
+      "step": 66300
+    },
+    {
+      "epoch": 0.6635,
+      "grad_norm": 39.52790069580078,
+      "learning_rate": 7.347320000000001e-07,
+      "loss": 0.332,
+      "step": 66350
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 12.687024116516113,
+      "learning_rate": 7.34532e-07,
+      "loss": 0.4187,
+      "step": 66400
+    },
+    {
+      "epoch": 0.6645,
+      "grad_norm": 79.7389144897461,
+      "learning_rate": 7.343319999999999e-07,
+      "loss": 0.3296,
+      "step": 66450
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 99.39762115478516,
+      "learning_rate": 7.34132e-07,
+      "loss": 0.44,
+      "step": 66500
+    },
+    {
+      "epoch": 0.6655,
+      "grad_norm": 1.2789992094039917,
+      "learning_rate": 7.33932e-07,
+      "loss": 0.4709,
+      "step": 66550
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 102.41425323486328,
+      "learning_rate": 7.33732e-07,
+      "loss": 0.4785,
+      "step": 66600
+    },
+    {
+      "epoch": 0.6665,
+      "grad_norm": 1.850731611251831,
+      "learning_rate": 7.33532e-07,
+      "loss": 0.5505,
+      "step": 66650
+    },
+    {
+      "epoch": 0.667,
+      "grad_norm": 6.710739612579346,
+      "learning_rate": 7.33332e-07,
+      "loss": 0.4497,
+      "step": 66700
+    },
+    {
+      "epoch": 0.6675,
+      "grad_norm": 27.49475860595703,
+      "learning_rate": 7.33132e-07,
+      "loss": 0.4216,
+      "step": 66750
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 109.29817962646484,
+      "learning_rate": 7.32932e-07,
+      "loss": 0.5599,
+      "step": 66800
+    },
+    {
+      "epoch": 0.6685,
+      "grad_norm": 2.05843186378479,
+      "learning_rate": 7.32732e-07,
+      "loss": 0.4372,
+      "step": 66850
+    },
+    {
+      "epoch": 0.669,
+      "grad_norm": 19.476158142089844,
+      "learning_rate": 7.325319999999999e-07,
+      "loss": 0.573,
+      "step": 66900
+    },
+    {
+      "epoch": 0.6695,
+      "grad_norm": 78.01306915283203,
+      "learning_rate": 7.32332e-07,
+      "loss": 0.3769,
+      "step": 66950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 64.33741760253906,
+      "learning_rate": 7.32132e-07,
+      "loss": 0.4052,
+      "step": 67000
+    },
+    {
+      "epoch": 0.6705,
+      "grad_norm": 2.576007843017578,
+      "learning_rate": 7.31932e-07,
+      "loss": 0.5676,
+      "step": 67050
+    },
+    {
+      "epoch": 0.671,
+      "grad_norm": 22.989377975463867,
+      "learning_rate": 7.317320000000001e-07,
+      "loss": 0.511,
+      "step": 67100
+    },
+    {
+      "epoch": 0.6715,
+      "grad_norm": 0.35649439692497253,
+      "learning_rate": 7.315319999999999e-07,
+      "loss": 0.5617,
+      "step": 67150
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 95.40701293945312,
+      "learning_rate": 7.313319999999999e-07,
+      "loss": 0.5461,
+      "step": 67200
+    },
+    {
+      "epoch": 0.6725,
+      "grad_norm": 80.1086196899414,
+      "learning_rate": 7.31132e-07,
+      "loss": 0.6098,
+      "step": 67250
+    },
+    {
+      "epoch": 0.673,
+      "grad_norm": 119.30028533935547,
+      "learning_rate": 7.30932e-07,
+      "loss": 0.4937,
+      "step": 67300
+    },
+    {
+      "epoch": 0.6735,
+      "grad_norm": 38.92840576171875,
+      "learning_rate": 7.30732e-07,
+      "loss": 0.4933,
+      "step": 67350
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 113.92001342773438,
+      "learning_rate": 7.30532e-07,
+      "loss": 0.421,
+      "step": 67400
+    },
+    {
+      "epoch": 0.6745,
+      "grad_norm": 0.898191511631012,
+      "learning_rate": 7.30332e-07,
+      "loss": 0.4647,
+      "step": 67450
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 62.35942077636719,
+      "learning_rate": 7.301319999999999e-07,
+      "loss": 0.5297,
+      "step": 67500
+    },
+    {
+      "epoch": 0.6755,
+      "grad_norm": 10.209319114685059,
+      "learning_rate": 7.29932e-07,
+      "loss": 0.4762,
+      "step": 67550
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 84.57727813720703,
+      "learning_rate": 7.29732e-07,
+      "loss": 0.6114,
+      "step": 67600
+    },
+    {
+      "epoch": 0.6765,
+      "grad_norm": 60.19552230834961,
+      "learning_rate": 7.295319999999999e-07,
+      "loss": 0.499,
+      "step": 67650
+    },
+    {
+      "epoch": 0.677,
+      "grad_norm": 12.361583709716797,
+      "learning_rate": 7.29332e-07,
+      "loss": 0.3799,
+      "step": 67700
+    },
+    {
+      "epoch": 0.6775,
+      "grad_norm": 72.01377868652344,
+      "learning_rate": 7.29132e-07,
+      "loss": 0.505,
+      "step": 67750
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 66.44666290283203,
+      "learning_rate": 7.28932e-07,
+      "loss": 0.6133,
+      "step": 67800
+    },
+    {
+      "epoch": 0.6785,
+      "grad_norm": 50.07161331176758,
+      "learning_rate": 7.28732e-07,
+      "loss": 0.4261,
+      "step": 67850
+    },
+    {
+      "epoch": 0.679,
+      "grad_norm": 104.22811889648438,
+      "learning_rate": 7.285319999999999e-07,
+      "loss": 0.5914,
+      "step": 67900
+    },
+    {
+      "epoch": 0.6795,
+      "grad_norm": 6.429470062255859,
+      "learning_rate": 7.283319999999999e-07,
+      "loss": 0.345,
+      "step": 67950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 65.51649475097656,
+      "learning_rate": 7.28132e-07,
+      "loss": 0.4633,
+      "step": 68000
+    },
+    {
+      "epoch": 0.6805,
+      "grad_norm": 2.1752583980560303,
+      "learning_rate": 7.27932e-07,
+      "loss": 0.4383,
+      "step": 68050
+    },
+    {
+      "epoch": 0.681,
+      "grad_norm": 22.099319458007812,
+      "learning_rate": 7.27732e-07,
+      "loss": 0.5433,
+      "step": 68100
+    },
+    {
+      "epoch": 0.6815,
+      "grad_norm": 117.40200805664062,
+      "learning_rate": 7.27532e-07,
+      "loss": 0.5817,
+      "step": 68150
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 57.13771057128906,
+      "learning_rate": 7.273319999999999e-07,
+      "loss": 0.4602,
+      "step": 68200
+    },
+    {
+      "epoch": 0.6825,
+      "grad_norm": 16.235754013061523,
+      "learning_rate": 7.271319999999999e-07,
+      "loss": 0.6009,
+      "step": 68250
+    },
+    {
+      "epoch": 0.683,
+      "grad_norm": 6.888073921203613,
+      "learning_rate": 7.26932e-07,
+      "loss": 0.4454,
+      "step": 68300
+    },
+    {
+      "epoch": 0.6835,
+      "grad_norm": 4.0116705894470215,
+      "learning_rate": 7.26732e-07,
+      "loss": 0.3861,
+      "step": 68350
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 36.852413177490234,
+      "learning_rate": 7.265319999999999e-07,
+      "loss": 0.6036,
+      "step": 68400
+    },
+    {
+      "epoch": 0.6845,
+      "grad_norm": 23.3376522064209,
+      "learning_rate": 7.26332e-07,
+      "loss": 0.4487,
+      "step": 68450
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 12.914100646972656,
+      "learning_rate": 7.26132e-07,
+      "loss": 0.5701,
+      "step": 68500
+    },
+    {
+      "epoch": 0.6855,
+      "grad_norm": 39.22294616699219,
+      "learning_rate": 7.259319999999999e-07,
+      "loss": 0.6363,
+      "step": 68550
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 38.24320602416992,
+      "learning_rate": 7.25732e-07,
+      "loss": 0.3829,
+      "step": 68600
+    },
+    {
+      "epoch": 0.6865,
+      "grad_norm": 28.683549880981445,
+      "learning_rate": 7.255319999999999e-07,
+      "loss": 0.444,
+      "step": 68650
+    },
+    {
+      "epoch": 0.687,
+      "grad_norm": 14.346872329711914,
+      "learning_rate": 7.253319999999999e-07,
+      "loss": 0.5185,
+      "step": 68700
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.12592576444149017,
+      "learning_rate": 7.25132e-07,
+      "loss": 0.512,
+      "step": 68750
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 21.978870391845703,
+      "learning_rate": 7.24932e-07,
+      "loss": 0.4442,
+      "step": 68800
+    },
+    {
+      "epoch": 0.6885,
+      "grad_norm": 0.3890920579433441,
+      "learning_rate": 7.247320000000001e-07,
+      "loss": 0.4323,
+      "step": 68850
+    },
+    {
+      "epoch": 0.689,
+      "grad_norm": 46.296112060546875,
+      "learning_rate": 7.245319999999999e-07,
+      "loss": 0.5979,
+      "step": 68900
+    },
+    {
+      "epoch": 0.6895,
+      "grad_norm": 3.0906097888946533,
+      "learning_rate": 7.243319999999999e-07,
+      "loss": 0.4466,
+      "step": 68950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 113.38265991210938,
+      "learning_rate": 7.24132e-07,
+      "loss": 0.4887,
+      "step": 69000
+    },
+    {
+      "epoch": 0.6905,
+      "grad_norm": 81.5102310180664,
+      "learning_rate": 7.23932e-07,
+      "loss": 0.6562,
+      "step": 69050
+    },
+    {
+      "epoch": 0.691,
+      "grad_norm": 0.32676833868026733,
+      "learning_rate": 7.23732e-07,
+      "loss": 0.4413,
+      "step": 69100
+    },
+    {
+      "epoch": 0.6915,
+      "grad_norm": 2.3341333866119385,
+      "learning_rate": 7.23532e-07,
+      "loss": 0.4187,
+      "step": 69150
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 51.32745361328125,
+      "learning_rate": 7.23332e-07,
+      "loss": 0.5124,
+      "step": 69200
+    },
+    {
+      "epoch": 0.6925,
+      "grad_norm": 106.96540069580078,
+      "learning_rate": 7.231319999999999e-07,
+      "loss": 0.4508,
+      "step": 69250
+    },
+    {
+      "epoch": 0.693,
+      "grad_norm": 28.311433792114258,
+      "learning_rate": 7.22932e-07,
+      "loss": 0.4263,
+      "step": 69300
+    },
+    {
+      "epoch": 0.6935,
+      "grad_norm": 31.820341110229492,
+      "learning_rate": 7.22732e-07,
+      "loss": 0.5164,
+      "step": 69350
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 56.47795867919922,
+      "learning_rate": 7.225319999999999e-07,
+      "loss": 0.535,
+      "step": 69400
+    },
+    {
+      "epoch": 0.6945,
+      "grad_norm": 10.806572914123535,
+      "learning_rate": 7.22332e-07,
+      "loss": 0.4787,
+      "step": 69450
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 41.71363067626953,
+      "learning_rate": 7.22132e-07,
+      "loss": 0.5698,
+      "step": 69500
+    },
+    {
+      "epoch": 0.6955,
+      "grad_norm": 14.216658592224121,
+      "learning_rate": 7.21932e-07,
+      "loss": 0.4295,
+      "step": 69550
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 73.08179473876953,
+      "learning_rate": 7.21732e-07,
+      "loss": 0.4647,
+      "step": 69600
+    },
+    {
+      "epoch": 0.6965,
+      "grad_norm": 32.81566619873047,
+      "learning_rate": 7.215319999999999e-07,
+      "loss": 0.4516,
+      "step": 69650
+    },
+    {
+      "epoch": 0.697,
+      "grad_norm": 74.85893249511719,
+      "learning_rate": 7.213319999999999e-07,
+      "loss": 0.5258,
+      "step": 69700
+    },
+    {
+      "epoch": 0.6975,
+      "grad_norm": 15.480995178222656,
+      "learning_rate": 7.21132e-07,
+      "loss": 0.48,
+      "step": 69750
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 36.054656982421875,
+      "learning_rate": 7.20932e-07,
+      "loss": 0.5223,
+      "step": 69800
+    },
+    {
+      "epoch": 0.6985,
+      "grad_norm": 6.590498447418213,
+      "learning_rate": 7.20732e-07,
+      "loss": 0.6518,
+      "step": 69850
+    },
+    {
+      "epoch": 0.699,
+      "grad_norm": 34.523651123046875,
+      "learning_rate": 7.20532e-07,
+      "loss": 0.6054,
+      "step": 69900
+    },
+    {
+      "epoch": 0.6995,
+      "grad_norm": 413.315185546875,
+      "learning_rate": 7.203319999999999e-07,
+      "loss": 0.3861,
+      "step": 69950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 28.01694679260254,
+      "learning_rate": 7.201319999999999e-07,
+      "loss": 0.5878,
+      "step": 70000
+    },
+    {
+      "epoch": 0.7005,
+      "grad_norm": 4.377991676330566,
+      "learning_rate": 7.19932e-07,
+      "loss": 0.5515,
+      "step": 70050
+    },
+    {
+      "epoch": 0.701,
+      "grad_norm": 55.94136428833008,
+      "learning_rate": 7.19732e-07,
+      "loss": 0.4695,
+      "step": 70100
+    },
+    {
+      "epoch": 0.7015,
+      "grad_norm": 68.08269500732422,
+      "learning_rate": 7.195359999999999e-07,
+      "loss": 0.5501,
+      "step": 70150
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 13.8864107131958,
+      "learning_rate": 7.19336e-07,
+      "loss": 0.4736,
+      "step": 70200
+    },
+    {
+      "epoch": 0.7025,
+      "grad_norm": 12.973597526550293,
+      "learning_rate": 7.19136e-07,
+      "loss": 0.6711,
+      "step": 70250
+    },
+    {
+      "epoch": 0.703,
+      "grad_norm": 59.53081512451172,
+      "learning_rate": 7.189359999999999e-07,
+      "loss": 0.6633,
+      "step": 70300
+    },
+    {
+      "epoch": 0.7035,
+      "grad_norm": 33.69902420043945,
+      "learning_rate": 7.18736e-07,
+      "loss": 0.5304,
+      "step": 70350
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 4.359379768371582,
+      "learning_rate": 7.185359999999999e-07,
+      "loss": 0.5243,
+      "step": 70400
+    },
+    {
+      "epoch": 0.7045,
+      "grad_norm": 42.60222625732422,
+      "learning_rate": 7.183359999999999e-07,
+      "loss": 0.4602,
+      "step": 70450
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 1.3176884651184082,
+      "learning_rate": 7.18136e-07,
+      "loss": 0.4459,
+      "step": 70500
+    },
+    {
+      "epoch": 0.7055,
+      "grad_norm": 3.021962881088257,
+      "learning_rate": 7.17936e-07,
+      "loss": 0.4491,
+      "step": 70550
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 274.6659851074219,
+      "learning_rate": 7.17736e-07,
+      "loss": 0.5348,
+      "step": 70600
+    },
+    {
+      "epoch": 0.7065,
+      "grad_norm": 59.16705322265625,
+      "learning_rate": 7.17536e-07,
+      "loss": 0.4527,
+      "step": 70650
+    },
+    {
+      "epoch": 0.707,
+      "grad_norm": 19.327486038208008,
+      "learning_rate": 7.173359999999999e-07,
+      "loss": 0.6739,
+      "step": 70700
+    },
+    {
+      "epoch": 0.7075,
+      "grad_norm": 10.360088348388672,
+      "learning_rate": 7.171359999999999e-07,
+      "loss": 0.621,
+      "step": 70750
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 65.7010726928711,
+      "learning_rate": 7.16936e-07,
+      "loss": 0.3637,
+      "step": 70800
+    },
+    {
+      "epoch": 0.7085,
+      "grad_norm": 7.891746997833252,
+      "learning_rate": 7.16736e-07,
+      "loss": 0.4372,
+      "step": 70850
+    },
+    {
+      "epoch": 0.709,
+      "grad_norm": 46.58807373046875,
+      "learning_rate": 7.165359999999999e-07,
+      "loss": 0.4355,
+      "step": 70900
+    },
+    {
+      "epoch": 0.7095,
+      "grad_norm": 82.19285583496094,
+      "learning_rate": 7.16336e-07,
+      "loss": 0.3844,
+      "step": 70950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 44.19703674316406,
+      "learning_rate": 7.16136e-07,
+      "loss": 0.4022,
+      "step": 71000
+    },
+    {
+      "epoch": 0.7105,
+      "grad_norm": 64.81764221191406,
+      "learning_rate": 7.159359999999999e-07,
+      "loss": 0.4597,
+      "step": 71050
+    },
+    {
+      "epoch": 0.711,
+      "grad_norm": 4.582440376281738,
+      "learning_rate": 7.15736e-07,
+      "loss": 0.3086,
+      "step": 71100
+    },
+    {
+      "epoch": 0.7115,
+      "grad_norm": 95.07144165039062,
+      "learning_rate": 7.155359999999999e-07,
+      "loss": 0.4794,
+      "step": 71150
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 47.082794189453125,
+      "learning_rate": 7.153359999999999e-07,
+      "loss": 0.5136,
+      "step": 71200
+    },
+    {
+      "epoch": 0.7125,
+      "grad_norm": 58.89522171020508,
+      "learning_rate": 7.15136e-07,
+      "loss": 0.4111,
+      "step": 71250
+    },
+    {
+      "epoch": 0.713,
+      "grad_norm": 76.09228515625,
+      "learning_rate": 7.14936e-07,
+      "loss": 0.493,
+      "step": 71300
+    },
+    {
+      "epoch": 0.7135,
+      "grad_norm": 7.7800211906433105,
+      "learning_rate": 7.147360000000001e-07,
+      "loss": 0.3391,
+      "step": 71350
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 24.84239959716797,
+      "learning_rate": 7.145359999999999e-07,
+      "loss": 0.5557,
+      "step": 71400
+    },
+    {
+      "epoch": 0.7145,
+      "grad_norm": 1.11128568649292,
+      "learning_rate": 7.143359999999999e-07,
+      "loss": 0.4902,
+      "step": 71450
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 67.38085174560547,
+      "learning_rate": 7.14136e-07,
+      "loss": 0.5157,
+      "step": 71500
+    },
+    {
+      "epoch": 0.7155,
+      "grad_norm": 58.82173538208008,
+      "learning_rate": 7.13936e-07,
+      "loss": 0.4311,
+      "step": 71550
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 10.34243106842041,
+      "learning_rate": 7.13736e-07,
+      "loss": 0.5631,
+      "step": 71600
+    },
+    {
+      "epoch": 0.7165,
+      "grad_norm": 31.23766326904297,
+      "learning_rate": 7.13536e-07,
+      "loss": 0.4615,
+      "step": 71650
+    },
+    {
+      "epoch": 0.717,
+      "grad_norm": 90.4206314086914,
+      "learning_rate": 7.13336e-07,
+      "loss": 0.6139,
+      "step": 71700
+    },
+    {
+      "epoch": 0.7175,
+      "grad_norm": 17.178178787231445,
+      "learning_rate": 7.131359999999999e-07,
+      "loss": 0.5126,
+      "step": 71750
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 60.4676399230957,
+      "learning_rate": 7.12936e-07,
+      "loss": 0.6,
+      "step": 71800
+    },
+    {
+      "epoch": 0.7185,
+      "grad_norm": 1.8041740655899048,
+      "learning_rate": 7.12736e-07,
+      "loss": 0.4853,
+      "step": 71850
+    },
+    {
+      "epoch": 0.719,
+      "grad_norm": 60.83064270019531,
+      "learning_rate": 7.125359999999999e-07,
+      "loss": 0.6092,
+      "step": 71900
+    },
+    {
+      "epoch": 0.7195,
+      "grad_norm": 59.2396354675293,
+      "learning_rate": 7.12336e-07,
+      "loss": 0.5124,
+      "step": 71950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 66.30610656738281,
+      "learning_rate": 7.12136e-07,
+      "loss": 0.5062,
+      "step": 72000
+    },
+    {
+      "epoch": 0.7205,
+      "grad_norm": 116.58683013916016,
+      "learning_rate": 7.11936e-07,
+      "loss": 0.561,
+      "step": 72050
+    },
+    {
+      "epoch": 0.721,
+      "grad_norm": 64.6193618774414,
+      "learning_rate": 7.11736e-07,
+      "loss": 0.5903,
+      "step": 72100
+    },
+    {
+      "epoch": 0.7215,
+      "grad_norm": 29.871625900268555,
+      "learning_rate": 7.115359999999999e-07,
+      "loss": 0.3967,
+      "step": 72150
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 48.294960021972656,
+      "learning_rate": 7.113359999999999e-07,
+      "loss": 0.4418,
+      "step": 72200
+    },
+    {
+      "epoch": 0.7225,
+      "grad_norm": 1.5724999904632568,
+      "learning_rate": 7.11136e-07,
+      "loss": 0.5528,
+      "step": 72250
+    },
+    {
+      "epoch": 0.723,
+      "grad_norm": 10.528818130493164,
+      "learning_rate": 7.10936e-07,
+      "loss": 0.3585,
+      "step": 72300
+    },
+    {
+      "epoch": 0.7235,
+      "grad_norm": 26.029348373413086,
+      "learning_rate": 7.10736e-07,
+      "loss": 0.5177,
+      "step": 72350
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 92.28126525878906,
+      "learning_rate": 7.10536e-07,
+      "loss": 0.3724,
+      "step": 72400
+    },
+    {
+      "epoch": 0.7245,
+      "grad_norm": 14.755386352539062,
+      "learning_rate": 7.103359999999999e-07,
+      "loss": 0.4839,
+      "step": 72450
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 32.792755126953125,
+      "learning_rate": 7.101359999999999e-07,
+      "loss": 0.5051,
+      "step": 72500
+    },
+    {
+      "epoch": 0.7255,
+      "grad_norm": 44.9954948425293,
+      "learning_rate": 7.09936e-07,
+      "loss": 0.572,
+      "step": 72550
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 61.472633361816406,
+      "learning_rate": 7.09736e-07,
+      "loss": 0.3723,
+      "step": 72600
+    },
+    {
+      "epoch": 0.7265,
+      "grad_norm": 56.659156799316406,
+      "learning_rate": 7.095359999999999e-07,
+      "loss": 0.5166,
+      "step": 72650
+    },
+    {
+      "epoch": 0.727,
+      "grad_norm": 66.07804107666016,
+      "learning_rate": 7.09336e-07,
+      "loss": 0.5227,
+      "step": 72700
+    },
+    {
+      "epoch": 0.7275,
+      "grad_norm": 18.027997970581055,
+      "learning_rate": 7.09136e-07,
+      "loss": 0.5235,
+      "step": 72750
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 71.68598937988281,
+      "learning_rate": 7.089359999999999e-07,
+      "loss": 0.4355,
+      "step": 72800
+    },
+    {
+      "epoch": 0.7285,
+      "grad_norm": 0.8782522082328796,
+      "learning_rate": 7.08736e-07,
+      "loss": 0.4416,
+      "step": 72850
+    },
+    {
+      "epoch": 0.729,
+      "grad_norm": 66.14490509033203,
+      "learning_rate": 7.08536e-07,
+      "loss": 0.4943,
+      "step": 72900
+    },
+    {
+      "epoch": 0.7295,
+      "grad_norm": 17.617177963256836,
+      "learning_rate": 7.083359999999999e-07,
+      "loss": 0.597,
+      "step": 72950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 95.89069366455078,
+      "learning_rate": 7.08136e-07,
+      "loss": 0.5914,
+      "step": 73000
+    },
+    {
+      "epoch": 0.7305,
+      "grad_norm": 6.234703063964844,
+      "learning_rate": 7.07936e-07,
+      "loss": 0.432,
+      "step": 73050
+    },
+    {
+      "epoch": 0.731,
+      "grad_norm": 81.79715728759766,
+      "learning_rate": 7.07736e-07,
+      "loss": 0.5353,
+      "step": 73100
+    },
+    {
+      "epoch": 0.7315,
+      "grad_norm": 68.71401977539062,
+      "learning_rate": 7.07536e-07,
+      "loss": 0.4773,
+      "step": 73150
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 20.033000946044922,
+      "learning_rate": 7.073359999999999e-07,
+      "loss": 0.3961,
+      "step": 73200
+    },
+    {
+      "epoch": 0.7325,
+      "grad_norm": 0.8714961409568787,
+      "learning_rate": 7.071359999999999e-07,
+      "loss": 0.4119,
+      "step": 73250
+    },
+    {
+      "epoch": 0.733,
+      "grad_norm": 67.67875671386719,
+      "learning_rate": 7.06936e-07,
+      "loss": 0.4616,
+      "step": 73300
+    },
+    {
+      "epoch": 0.7335,
+      "grad_norm": 170.6661376953125,
+      "learning_rate": 7.06736e-07,
+      "loss": 0.4623,
+      "step": 73350
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 22.065210342407227,
+      "learning_rate": 7.06536e-07,
+      "loss": 0.5301,
+      "step": 73400
+    },
+    {
+      "epoch": 0.7345,
+      "grad_norm": 116.2767562866211,
+      "learning_rate": 7.06336e-07,
+      "loss": 0.6367,
+      "step": 73450
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 118.61943817138672,
+      "learning_rate": 7.061359999999999e-07,
+      "loss": 0.4786,
+      "step": 73500
+    },
+    {
+      "epoch": 0.7355,
+      "grad_norm": 10.633493423461914,
+      "learning_rate": 7.05936e-07,
+      "loss": 0.5042,
+      "step": 73550
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 15.222790718078613,
+      "learning_rate": 7.05736e-07,
+      "loss": 0.5449,
+      "step": 73600
+    },
+    {
+      "epoch": 0.7365,
+      "grad_norm": 4.541469097137451,
+      "learning_rate": 7.05536e-07,
+      "loss": 0.4413,
+      "step": 73650
+    },
+    {
+      "epoch": 0.737,
+      "grad_norm": 50.30397033691406,
+      "learning_rate": 7.05336e-07,
+      "loss": 0.4317,
+      "step": 73700
+    },
+    {
+      "epoch": 0.7375,
+      "grad_norm": 4.888488292694092,
+      "learning_rate": 7.05136e-07,
+      "loss": 0.4241,
+      "step": 73750
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 46.71607971191406,
+      "learning_rate": 7.04936e-07,
+      "loss": 0.4109,
+      "step": 73800
+    },
+    {
+      "epoch": 0.7385,
+      "grad_norm": 139.61004638671875,
+      "learning_rate": 7.04736e-07,
+      "loss": 0.5672,
+      "step": 73850
+    },
+    {
+      "epoch": 0.739,
+      "grad_norm": 96.24576568603516,
+      "learning_rate": 7.04536e-07,
+      "loss": 0.5308,
+      "step": 73900
+    },
+    {
+      "epoch": 0.7395,
+      "grad_norm": 3.3251993656158447,
+      "learning_rate": 7.043359999999999e-07,
+      "loss": 0.5206,
+      "step": 73950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 64.65046691894531,
+      "learning_rate": 7.04136e-07,
+      "loss": 0.3965,
+      "step": 74000
+    },
+    {
+      "epoch": 0.7405,
+      "grad_norm": 79.69112396240234,
+      "learning_rate": 7.03936e-07,
+      "loss": 0.571,
+      "step": 74050
+    },
+    {
+      "epoch": 0.741,
+      "grad_norm": 86.38766479492188,
+      "learning_rate": 7.03736e-07,
+      "loss": 0.6437,
+      "step": 74100
+    },
+    {
+      "epoch": 0.7415,
+      "grad_norm": 43.816471099853516,
+      "learning_rate": 7.035360000000001e-07,
+      "loss": 0.4205,
+      "step": 74150
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 41.845672607421875,
+      "learning_rate": 7.033359999999999e-07,
+      "loss": 0.3473,
+      "step": 74200
+    },
+    {
+      "epoch": 0.7425,
+      "grad_norm": 0.8183789849281311,
+      "learning_rate": 7.031359999999999e-07,
+      "loss": 0.4469,
+      "step": 74250
+    },
+    {
+      "epoch": 0.743,
+      "grad_norm": 3.786691427230835,
+      "learning_rate": 7.02936e-07,
+      "loss": 0.5015,
+      "step": 74300
+    },
+    {
+      "epoch": 0.7435,
+      "grad_norm": 78.48735046386719,
+      "learning_rate": 7.02736e-07,
+      "loss": 0.624,
+      "step": 74350
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 104.10353088378906,
+      "learning_rate": 7.02536e-07,
+      "loss": 0.4875,
+      "step": 74400
+    },
+    {
+      "epoch": 0.7445,
+      "grad_norm": 6.0870819091796875,
+      "learning_rate": 7.02336e-07,
+      "loss": 0.4004,
+      "step": 74450
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 13.105294227600098,
+      "learning_rate": 7.02136e-07,
+      "loss": 0.4386,
+      "step": 74500
+    },
+    {
+      "epoch": 0.7455,
+      "grad_norm": 51.39162826538086,
+      "learning_rate": 7.019359999999999e-07,
+      "loss": 0.4182,
+      "step": 74550
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 91.84526824951172,
+      "learning_rate": 7.01736e-07,
+      "loss": 0.5552,
+      "step": 74600
+    },
+    {
+      "epoch": 0.7465,
+      "grad_norm": 9.574994087219238,
+      "learning_rate": 7.01536e-07,
+      "loss": 0.4372,
+      "step": 74650
+    },
+    {
+      "epoch": 0.747,
+      "grad_norm": 75.70382690429688,
+      "learning_rate": 7.013359999999999e-07,
+      "loss": 0.5477,
+      "step": 74700
+    },
+    {
+      "epoch": 0.7475,
+      "grad_norm": 59.122093200683594,
+      "learning_rate": 7.01136e-07,
+      "loss": 0.5786,
+      "step": 74750
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 92.53333282470703,
+      "learning_rate": 7.00936e-07,
+      "loss": 0.5653,
+      "step": 74800
+    },
+    {
+      "epoch": 0.7485,
+      "grad_norm": 93.97532653808594,
+      "learning_rate": 7.00736e-07,
+      "loss": 0.3929,
+      "step": 74850
+    },
+    {
+      "epoch": 0.749,
+      "grad_norm": 1.357946515083313,
+      "learning_rate": 7.00536e-07,
+      "loss": 0.3806,
+      "step": 74900
+    },
+    {
+      "epoch": 0.7495,
+      "grad_norm": 1.4514474868774414,
+      "learning_rate": 7.003359999999999e-07,
+      "loss": 0.4542,
+      "step": 74950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 60.96038055419922,
+      "learning_rate": 7.001359999999999e-07,
+      "loss": 0.4032,
+      "step": 75000
+    },
+    {
+      "epoch": 0.7505,
+      "grad_norm": 54.15022659301758,
+      "learning_rate": 6.99936e-07,
+      "loss": 0.5132,
+      "step": 75050
+    },
+    {
+      "epoch": 0.751,
+      "grad_norm": 92.10535430908203,
+      "learning_rate": 6.99736e-07,
+      "loss": 0.5717,
+      "step": 75100
+    },
+    {
+      "epoch": 0.7515,
+      "grad_norm": 67.41671752929688,
+      "learning_rate": 6.99536e-07,
+      "loss": 0.4319,
+      "step": 75150
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 27.999591827392578,
+      "learning_rate": 6.99336e-07,
+      "loss": 0.5642,
+      "step": 75200
+    },
+    {
+      "epoch": 0.7525,
+      "grad_norm": 54.86029815673828,
+      "learning_rate": 6.991359999999999e-07,
+      "loss": 0.6475,
+      "step": 75250
+    },
+    {
+      "epoch": 0.753,
+      "grad_norm": 38.79415512084961,
+      "learning_rate": 6.989359999999999e-07,
+      "loss": 0.3514,
+      "step": 75300
+    },
+    {
+      "epoch": 0.7535,
+      "grad_norm": 2.452829360961914,
+      "learning_rate": 6.98736e-07,
+      "loss": 0.5446,
+      "step": 75350
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 118.6248550415039,
+      "learning_rate": 6.98536e-07,
+      "loss": 0.5,
+      "step": 75400
+    },
+    {
+      "epoch": 0.7545,
+      "grad_norm": 70.82340240478516,
+      "learning_rate": 6.983359999999999e-07,
+      "loss": 0.5008,
+      "step": 75450
+    },
+    {
+      "epoch": 0.755,
+      "grad_norm": 42.31960678100586,
+      "learning_rate": 6.9814e-07,
+      "loss": 0.4178,
+      "step": 75500
+    },
+    {
+      "epoch": 0.7555,
+      "grad_norm": 90.04888916015625,
+      "learning_rate": 6.9794e-07,
+      "loss": 0.5661,
+      "step": 75550
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 21.198366165161133,
+      "learning_rate": 6.977399999999999e-07,
+      "loss": 0.4164,
+      "step": 75600
+    },
+    {
+      "epoch": 0.7565,
+      "grad_norm": 78.9493408203125,
+      "learning_rate": 6.9754e-07,
+      "loss": 0.5132,
+      "step": 75650
+    },
+    {
+      "epoch": 0.757,
+      "grad_norm": 103.06893920898438,
+      "learning_rate": 6.973399999999999e-07,
+      "loss": 0.5996,
+      "step": 75700
+    },
+    {
+      "epoch": 0.7575,
+      "grad_norm": 4.515233993530273,
+      "learning_rate": 6.971399999999999e-07,
+      "loss": 0.4079,
+      "step": 75750
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 68.41087341308594,
+      "learning_rate": 6.9694e-07,
+      "loss": 0.557,
+      "step": 75800
+    },
+    {
+      "epoch": 0.7585,
+      "grad_norm": 4.536024570465088,
+      "learning_rate": 6.9674e-07,
+      "loss": 0.4946,
+      "step": 75850
+    },
+    {
+      "epoch": 0.759,
+      "grad_norm": 43.321983337402344,
+      "learning_rate": 6.965400000000001e-07,
+      "loss": 0.4212,
+      "step": 75900
+    },
+    {
+      "epoch": 0.7595,
+      "grad_norm": 57.4993896484375,
+      "learning_rate": 6.9634e-07,
+      "loss": 0.4643,
+      "step": 75950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 58.03748321533203,
+      "learning_rate": 6.961399999999999e-07,
+      "loss": 0.5878,
+      "step": 76000
+    },
+    {
+      "epoch": 0.7605,
+      "grad_norm": 5.033849239349365,
+      "learning_rate": 6.9594e-07,
+      "loss": 0.6593,
+      "step": 76050
+    },
+    {
+      "epoch": 0.761,
+      "grad_norm": 75.7629623413086,
+      "learning_rate": 6.9574e-07,
+      "loss": 0.4768,
+      "step": 76100
+    },
+    {
+      "epoch": 0.7615,
+      "grad_norm": 60.79175567626953,
+      "learning_rate": 6.9554e-07,
+      "loss": 0.5528,
+      "step": 76150
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 71.26956176757812,
+      "learning_rate": 6.9534e-07,
+      "loss": 0.587,
+      "step": 76200
+    },
+    {
+      "epoch": 0.7625,
+      "grad_norm": 3.231938362121582,
+      "learning_rate": 6.9514e-07,
+      "loss": 0.4488,
+      "step": 76250
+    },
+    {
+      "epoch": 0.763,
+      "grad_norm": 2.168018102645874,
+      "learning_rate": 6.9494e-07,
+      "loss": 0.4011,
+      "step": 76300
+    },
+    {
+      "epoch": 0.7635,
+      "grad_norm": 2.409412145614624,
+      "learning_rate": 6.9474e-07,
+      "loss": 0.5966,
+      "step": 76350
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 8.960734367370605,
+      "learning_rate": 6.9454e-07,
+      "loss": 0.4323,
+      "step": 76400
+    },
+    {
+      "epoch": 0.7645,
+      "grad_norm": 74.19822692871094,
+      "learning_rate": 6.943399999999999e-07,
+      "loss": 0.3678,
+      "step": 76450
+    },
+    {
+      "epoch": 0.765,
+      "grad_norm": 75.17100524902344,
+      "learning_rate": 6.9414e-07,
+      "loss": 0.4418,
+      "step": 76500
+    },
+    {
+      "epoch": 0.7655,
+      "grad_norm": 23.91132164001465,
+      "learning_rate": 6.9394e-07,
+      "loss": 0.4939,
+      "step": 76550
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 33.52251052856445,
+      "learning_rate": 6.9374e-07,
+      "loss": 0.4463,
+      "step": 76600
+    },
+    {
+      "epoch": 0.7665,
+      "grad_norm": 119.65557861328125,
+      "learning_rate": 6.935400000000001e-07,
+      "loss": 0.6833,
+      "step": 76650
+    },
+    {
+      "epoch": 0.767,
+      "grad_norm": 46.15104675292969,
+      "learning_rate": 6.933399999999999e-07,
+      "loss": 0.4756,
+      "step": 76700
+    },
+    {
+      "epoch": 0.7675,
+      "grad_norm": 1.8948994874954224,
+      "learning_rate": 6.931399999999999e-07,
+      "loss": 0.535,
+      "step": 76750
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 38.17501449584961,
+      "learning_rate": 6.9294e-07,
+      "loss": 0.428,
+      "step": 76800
+    },
+    {
+      "epoch": 0.7685,
+      "grad_norm": 65.1424789428711,
+      "learning_rate": 6.9274e-07,
+      "loss": 0.5542,
+      "step": 76850
+    },
+    {
+      "epoch": 0.769,
+      "grad_norm": 88.36298370361328,
+      "learning_rate": 6.9254e-07,
+      "loss": 0.3903,
+      "step": 76900
+    },
+    {
+      "epoch": 0.7695,
+      "grad_norm": 152.0764923095703,
+      "learning_rate": 6.9234e-07,
+      "loss": 0.4659,
+      "step": 76950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 47.97085189819336,
+      "learning_rate": 6.9214e-07,
+      "loss": 0.5331,
+      "step": 77000
+    },
+    {
+      "epoch": 0.7705,
+      "grad_norm": 91.40066528320312,
+      "learning_rate": 6.919399999999999e-07,
+      "loss": 0.5595,
+      "step": 77050
+    },
+    {
+      "epoch": 0.771,
+      "grad_norm": 8.514924049377441,
+      "learning_rate": 6.9174e-07,
+      "loss": 0.3657,
+      "step": 77100
+    },
+    {
+      "epoch": 0.7715,
+      "grad_norm": 24.74837303161621,
+      "learning_rate": 6.9154e-07,
+      "loss": 0.4459,
+      "step": 77150
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 87.70143127441406,
+      "learning_rate": 6.913399999999999e-07,
+      "loss": 0.4602,
+      "step": 77200
+    },
+    {
+      "epoch": 0.7725,
+      "grad_norm": 20.411006927490234,
+      "learning_rate": 6.9114e-07,
+      "loss": 0.4812,
+      "step": 77250
+    },
+    {
+      "epoch": 0.773,
+      "grad_norm": 74.98430633544922,
+      "learning_rate": 6.9094e-07,
+      "loss": 0.4783,
+      "step": 77300
+    },
+    {
+      "epoch": 0.7735,
+      "grad_norm": 15.123334884643555,
+      "learning_rate": 6.9074e-07,
+      "loss": 0.289,
+      "step": 77350
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 81.70301818847656,
+      "learning_rate": 6.9054e-07,
+      "loss": 0.5672,
+      "step": 77400
+    },
+    {
+      "epoch": 0.7745,
+      "grad_norm": 0.1056661456823349,
+      "learning_rate": 6.903399999999999e-07,
+      "loss": 0.3384,
+      "step": 77450
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 40.181251525878906,
+      "learning_rate": 6.901399999999999e-07,
+      "loss": 0.5383,
+      "step": 77500
+    },
+    {
+      "epoch": 0.7755,
+      "grad_norm": 77.13603973388672,
+      "learning_rate": 6.8994e-07,
+      "loss": 0.5013,
+      "step": 77550
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 12.230585098266602,
+      "learning_rate": 6.8974e-07,
+      "loss": 0.4305,
+      "step": 77600
+    },
+    {
+      "epoch": 0.7765,
+      "grad_norm": 130.13206481933594,
+      "learning_rate": 6.8954e-07,
+      "loss": 0.4474,
+      "step": 77650
+    },
+    {
+      "epoch": 0.777,
+      "grad_norm": 7.258871555328369,
+      "learning_rate": 6.8934e-07,
+      "loss": 0.5191,
+      "step": 77700
+    },
+    {
+      "epoch": 0.7775,
+      "grad_norm": 74.94843292236328,
+      "learning_rate": 6.891399999999999e-07,
+      "loss": 0.4981,
+      "step": 77750
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 37.118385314941406,
+      "learning_rate": 6.889399999999999e-07,
+      "loss": 0.5874,
+      "step": 77800
+    },
+    {
+      "epoch": 0.7785,
+      "grad_norm": 66.5859146118164,
+      "learning_rate": 6.8874e-07,
+      "loss": 0.4819,
+      "step": 77850
+    },
+    {
+      "epoch": 0.779,
+      "grad_norm": 44.33268356323242,
+      "learning_rate": 6.8854e-07,
+      "loss": 0.5725,
+      "step": 77900
+    },
+    {
+      "epoch": 0.7795,
+      "grad_norm": 23.16948890686035,
+      "learning_rate": 6.883399999999999e-07,
+      "loss": 0.4699,
+      "step": 77950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 84.33336639404297,
+      "learning_rate": 6.8814e-07,
+      "loss": 0.5684,
+      "step": 78000
+    },
+    {
+      "epoch": 0.7805,
+      "grad_norm": 52.04594039916992,
+      "learning_rate": 6.8794e-07,
+      "loss": 0.5498,
+      "step": 78050
+    },
+    {
+      "epoch": 0.781,
+      "grad_norm": 9.365368843078613,
+      "learning_rate": 6.877399999999999e-07,
+      "loss": 0.4686,
+      "step": 78100
+    },
+    {
+      "epoch": 0.7815,
+      "grad_norm": 20.72178077697754,
+      "learning_rate": 6.8754e-07,
+      "loss": 0.5012,
+      "step": 78150
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 35.4743537902832,
+      "learning_rate": 6.873399999999999e-07,
+      "loss": 0.5656,
+      "step": 78200
+    },
+    {
+      "epoch": 0.7825,
+      "grad_norm": 58.698486328125,
+      "learning_rate": 6.871399999999999e-07,
+      "loss": 0.5637,
+      "step": 78250
+    },
+    {
+      "epoch": 0.783,
+      "grad_norm": 4.087533950805664,
+      "learning_rate": 6.8694e-07,
+      "loss": 0.4656,
+      "step": 78300
+    },
+    {
+      "epoch": 0.7835,
+      "grad_norm": 114.83053588867188,
+      "learning_rate": 6.8674e-07,
+      "loss": 0.5119,
+      "step": 78350
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 22.25564193725586,
+      "learning_rate": 6.865400000000001e-07,
+      "loss": 0.4064,
+      "step": 78400
+    },
+    {
+      "epoch": 0.7845,
+      "grad_norm": 25.038494110107422,
+      "learning_rate": 6.863399999999999e-07,
+      "loss": 0.4126,
+      "step": 78450
+    },
+    {
+      "epoch": 0.785,
+      "grad_norm": 77.9610824584961,
+      "learning_rate": 6.861399999999999e-07,
+      "loss": 0.4798,
+      "step": 78500
+    },
+    {
+      "epoch": 0.7855,
+      "grad_norm": 48.64069366455078,
+      "learning_rate": 6.8594e-07,
+      "loss": 0.4212,
+      "step": 78550
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 33.35960006713867,
+      "learning_rate": 6.8574e-07,
+      "loss": 0.4711,
+      "step": 78600
+    },
+    {
+      "epoch": 0.7865,
+      "grad_norm": 8.562115669250488,
+      "learning_rate": 6.8554e-07,
+      "loss": 0.5208,
+      "step": 78650
+    },
+    {
+      "epoch": 0.787,
+      "grad_norm": 84.58731079101562,
+      "learning_rate": 6.8534e-07,
+      "loss": 0.5262,
+      "step": 78700
+    },
+    {
+      "epoch": 0.7875,
+      "grad_norm": 35.85997009277344,
+      "learning_rate": 6.8514e-07,
+      "loss": 0.6896,
+      "step": 78750
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 84.28465270996094,
+      "learning_rate": 6.849399999999999e-07,
+      "loss": 0.6728,
+      "step": 78800
+    },
+    {
+      "epoch": 0.7885,
+      "grad_norm": 105.6515121459961,
+      "learning_rate": 6.8474e-07,
+      "loss": 0.5276,
+      "step": 78850
+    },
+    {
+      "epoch": 0.789,
+      "grad_norm": 74.69842529296875,
+      "learning_rate": 6.8454e-07,
+      "loss": 0.5207,
+      "step": 78900
+    },
+    {
+      "epoch": 0.7895,
+      "grad_norm": 60.617088317871094,
+      "learning_rate": 6.843399999999999e-07,
+      "loss": 0.4344,
+      "step": 78950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 64.80194854736328,
+      "learning_rate": 6.8414e-07,
+      "loss": 0.3582,
+      "step": 79000
+    },
+    {
+      "epoch": 0.7905,
+      "grad_norm": 9.421799659729004,
+      "learning_rate": 6.8394e-07,
+      "loss": 0.3847,
+      "step": 79050
+    },
+    {
+      "epoch": 0.791,
+      "grad_norm": 27.81099510192871,
+      "learning_rate": 6.8374e-07,
+      "loss": 0.5735,
+      "step": 79100
+    },
+    {
+      "epoch": 0.7915,
+      "grad_norm": 11.743189811706543,
+      "learning_rate": 6.8354e-07,
+      "loss": 0.3546,
+      "step": 79150
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 2.8396387100219727,
+      "learning_rate": 6.833399999999999e-07,
+      "loss": 0.3229,
+      "step": 79200
+    },
+    {
+      "epoch": 0.7925,
+      "grad_norm": 0.9796394109725952,
+      "learning_rate": 6.831399999999999e-07,
+      "loss": 0.4796,
+      "step": 79250
+    },
+    {
+      "epoch": 0.793,
+      "grad_norm": 64.07000732421875,
+      "learning_rate": 6.8294e-07,
+      "loss": 0.4583,
+      "step": 79300
+    },
+    {
+      "epoch": 0.7935,
+      "grad_norm": 91.14384460449219,
+      "learning_rate": 6.8274e-07,
+      "loss": 0.5557,
+      "step": 79350
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 67.72103881835938,
+      "learning_rate": 6.8254e-07,
+      "loss": 0.4973,
+      "step": 79400
+    },
+    {
+      "epoch": 0.7945,
+      "grad_norm": 76.02549743652344,
+      "learning_rate": 6.8234e-07,
+      "loss": 0.4128,
+      "step": 79450
+    },
+    {
+      "epoch": 0.795,
+      "grad_norm": 14.491253852844238,
+      "learning_rate": 6.821399999999999e-07,
+      "loss": 0.5534,
+      "step": 79500
+    },
+    {
+      "epoch": 0.7955,
+      "grad_norm": 0.9683569073677063,
+      "learning_rate": 6.819399999999999e-07,
+      "loss": 0.6786,
+      "step": 79550
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 73.29056549072266,
+      "learning_rate": 6.8174e-07,
+      "loss": 0.4019,
+      "step": 79600
+    },
+    {
+      "epoch": 0.7965,
+      "grad_norm": 44.23388671875,
+      "learning_rate": 6.8154e-07,
+      "loss": 0.4544,
+      "step": 79650
+    },
+    {
+      "epoch": 0.797,
+      "grad_norm": 0.06450020521879196,
+      "learning_rate": 6.813439999999999e-07,
+      "loss": 0.3565,
+      "step": 79700
+    },
+    {
+      "epoch": 0.7975,
+      "grad_norm": 79.00003814697266,
+      "learning_rate": 6.81144e-07,
+      "loss": 0.5056,
+      "step": 79750
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 25.8265438079834,
+      "learning_rate": 6.80944e-07,
+      "loss": 0.4529,
+      "step": 79800
+    },
+    {
+      "epoch": 0.7985,
+      "grad_norm": 67.03202056884766,
+      "learning_rate": 6.807439999999999e-07,
+      "loss": 0.4031,
+      "step": 79850
+    },
+    {
+      "epoch": 0.799,
+      "grad_norm": 67.00475311279297,
+      "learning_rate": 6.80544e-07,
+      "loss": 0.4486,
+      "step": 79900
+    },
+    {
+      "epoch": 0.7995,
+      "grad_norm": 11.94206714630127,
+      "learning_rate": 6.803439999999999e-07,
+      "loss": 0.4968,
+      "step": 79950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 73.1527099609375,
+      "learning_rate": 6.801439999999999e-07,
+      "loss": 0.379,
+      "step": 80000
+    },
+    {
+      "epoch": 0.8005,
+      "grad_norm": 34.18701934814453,
+      "learning_rate": 6.79944e-07,
+      "loss": 0.4137,
+      "step": 80050
+    },
+    {
+      "epoch": 0.801,
+      "grad_norm": 64.62896728515625,
+      "learning_rate": 6.79744e-07,
+      "loss": 0.6624,
+      "step": 80100
+    },
+    {
+      "epoch": 0.8015,
+      "grad_norm": 52.73872375488281,
+      "learning_rate": 6.79544e-07,
+      "loss": 0.4575,
+      "step": 80150
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 76.4136734008789,
+      "learning_rate": 6.793439999999999e-07,
+      "loss": 0.5554,
+      "step": 80200
+    },
+    {
+      "epoch": 0.8025,
+      "grad_norm": 27.00578498840332,
+      "learning_rate": 6.791439999999999e-07,
+      "loss": 0.5405,
+      "step": 80250
+    },
+    {
+      "epoch": 0.803,
+      "grad_norm": 33.01068115234375,
+      "learning_rate": 6.789439999999999e-07,
+      "loss": 0.5504,
+      "step": 80300
+    },
+    {
+      "epoch": 0.8035,
+      "grad_norm": 0.1820012480020523,
+      "learning_rate": 6.78752e-07,
+      "loss": 0.5296,
+      "step": 80350
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 73.24763488769531,
+      "learning_rate": 6.78552e-07,
+      "loss": 0.4904,
+      "step": 80400
+    },
+    {
+      "epoch": 0.8045,
+      "grad_norm": 53.44038772583008,
+      "learning_rate": 6.783519999999999e-07,
+      "loss": 0.283,
+      "step": 80450
+    },
+    {
+      "epoch": 0.805,
+      "grad_norm": 9.23548412322998,
+      "learning_rate": 6.78152e-07,
+      "loss": 0.524,
+      "step": 80500
+    },
+    {
+      "epoch": 0.8055,
+      "grad_norm": 8.429581642150879,
+      "learning_rate": 6.77952e-07,
+      "loss": 0.3238,
+      "step": 80550
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 0.890905499458313,
+      "learning_rate": 6.777519999999999e-07,
+      "loss": 0.4394,
+      "step": 80600
+    },
+    {
+      "epoch": 0.8065,
+      "grad_norm": 61.318058013916016,
+      "learning_rate": 6.77552e-07,
+      "loss": 0.4028,
+      "step": 80650
+    },
+    {
+      "epoch": 0.807,
+      "grad_norm": 32.087276458740234,
+      "learning_rate": 6.773519999999999e-07,
+      "loss": 0.6014,
+      "step": 80700
+    },
+    {
+      "epoch": 0.8075,
+      "grad_norm": 14.302053451538086,
+      "learning_rate": 6.771519999999999e-07,
+      "loss": 0.5009,
+      "step": 80750
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 20.004087448120117,
+      "learning_rate": 6.76952e-07,
+      "loss": 0.5234,
+      "step": 80800
+    },
+    {
+      "epoch": 0.8085,
+      "grad_norm": 58.03341293334961,
+      "learning_rate": 6.76752e-07,
+      "loss": 0.5761,
+      "step": 80850
+    },
+    {
+      "epoch": 0.809,
+      "grad_norm": 42.07494354248047,
+      "learning_rate": 6.765560000000001e-07,
+      "loss": 0.4014,
+      "step": 80900
+    },
+    {
+      "epoch": 0.8095,
+      "grad_norm": 11.237667083740234,
+      "learning_rate": 6.763559999999999e-07,
+      "loss": 0.6565,
+      "step": 80950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 54.023685455322266,
+      "learning_rate": 6.761559999999999e-07,
+      "loss": 0.5733,
+      "step": 81000
+    },
+    {
+      "epoch": 0.8105,
+      "grad_norm": 26.939653396606445,
+      "learning_rate": 6.75956e-07,
+      "loss": 0.5418,
+      "step": 81050
+    },
+    {
+      "epoch": 0.811,
+      "grad_norm": 72.97735595703125,
+      "learning_rate": 6.75756e-07,
+      "loss": 0.6132,
+      "step": 81100
+    },
+    {
+      "epoch": 0.8115,
+      "grad_norm": 30.797290802001953,
+      "learning_rate": 6.75556e-07,
+      "loss": 0.4381,
+      "step": 81150
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 193.12718200683594,
+      "learning_rate": 6.75356e-07,
+      "loss": 0.5038,
+      "step": 81200
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 42.33784103393555,
+      "learning_rate": 6.75156e-07,
+      "loss": 0.5907,
+      "step": 81250
+    },
+    {
+      "epoch": 0.813,
+      "grad_norm": 60.777767181396484,
+      "learning_rate": 6.749559999999999e-07,
+      "loss": 0.5999,
+      "step": 81300
+    },
+    {
+      "epoch": 0.8135,
+      "grad_norm": 4.074952602386475,
+      "learning_rate": 6.74756e-07,
+      "loss": 0.5535,
+      "step": 81350
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 19.895477294921875,
+      "learning_rate": 6.74556e-07,
+      "loss": 0.4116,
+      "step": 81400
+    },
+    {
+      "epoch": 0.8145,
+      "grad_norm": 8.462628364562988,
+      "learning_rate": 6.743559999999999e-07,
+      "loss": 0.4759,
+      "step": 81450
+    },
+    {
+      "epoch": 0.815,
+      "grad_norm": 55.037052154541016,
+      "learning_rate": 6.74156e-07,
+      "loss": 0.5654,
+      "step": 81500
+    },
+    {
+      "epoch": 0.8155,
+      "grad_norm": 65.07530975341797,
+      "learning_rate": 6.73956e-07,
+      "loss": 0.5132,
+      "step": 81550
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 64.06502532958984,
+      "learning_rate": 6.73756e-07,
+      "loss": 0.5159,
+      "step": 81600
+    },
+    {
+      "epoch": 0.8165,
+      "grad_norm": 73.82300567626953,
+      "learning_rate": 6.73556e-07,
+      "loss": 0.5446,
+      "step": 81650
+    },
+    {
+      "epoch": 0.817,
+      "grad_norm": 40.699676513671875,
+      "learning_rate": 6.733559999999999e-07,
+      "loss": 0.4982,
+      "step": 81700
+    },
+    {
+      "epoch": 0.8175,
+      "grad_norm": 94.0560531616211,
+      "learning_rate": 6.731559999999999e-07,
+      "loss": 0.5136,
+      "step": 81750
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 77.07228088378906,
+      "learning_rate": 6.72956e-07,
+      "loss": 0.5026,
+      "step": 81800
+    },
+    {
+      "epoch": 0.8185,
+      "grad_norm": 54.43427658081055,
+      "learning_rate": 6.72756e-07,
+      "loss": 0.5811,
+      "step": 81850
+    },
+    {
+      "epoch": 0.819,
+      "grad_norm": 32.514156341552734,
+      "learning_rate": 6.72556e-07,
+      "loss": 0.4385,
+      "step": 81900
+    },
+    {
+      "epoch": 0.8195,
+      "grad_norm": 8.645065307617188,
+      "learning_rate": 6.72356e-07,
+      "loss": 0.3586,
+      "step": 81950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 42.03317642211914,
+      "learning_rate": 6.72156e-07,
+      "loss": 0.413,
+      "step": 82000
+    },
+    {
+      "epoch": 0.8205,
+      "grad_norm": 0.5245881676673889,
+      "learning_rate": 6.719559999999999e-07,
+      "loss": 0.4099,
+      "step": 82050
+    },
+    {
+      "epoch": 0.821,
+      "grad_norm": 66.3309326171875,
+      "learning_rate": 6.71756e-07,
+      "loss": 0.7605,
+      "step": 82100
+    },
+    {
+      "epoch": 0.8215,
+      "grad_norm": 26.570140838623047,
+      "learning_rate": 6.71556e-07,
+      "loss": 0.4468,
+      "step": 82150
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 71.70770263671875,
+      "learning_rate": 6.713559999999999e-07,
+      "loss": 0.3865,
+      "step": 82200
+    },
+    {
+      "epoch": 0.8225,
+      "grad_norm": 29.463645935058594,
+      "learning_rate": 6.71156e-07,
+      "loss": 0.5206,
+      "step": 82250
+    },
+    {
+      "epoch": 0.823,
+      "grad_norm": 61.58549880981445,
+      "learning_rate": 6.70956e-07,
+      "loss": 0.5122,
+      "step": 82300
+    },
+    {
+      "epoch": 0.8235,
+      "grad_norm": 52.86707305908203,
+      "learning_rate": 6.70756e-07,
+      "loss": 0.5216,
+      "step": 82350
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 67.04204559326172,
+      "learning_rate": 6.70556e-07,
+      "loss": 0.5536,
+      "step": 82400
+    },
+    {
+      "epoch": 0.8245,
+      "grad_norm": 1.1900335550308228,
+      "learning_rate": 6.703559999999999e-07,
+      "loss": 0.4011,
+      "step": 82450
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 13.605680465698242,
+      "learning_rate": 6.701559999999999e-07,
+      "loss": 0.412,
+      "step": 82500
+    },
+    {
+      "epoch": 0.8255,
+      "grad_norm": 68.13459777832031,
+      "learning_rate": 6.69956e-07,
+      "loss": 0.5592,
+      "step": 82550
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 41.138648986816406,
+      "learning_rate": 6.69756e-07,
+      "loss": 0.5622,
+      "step": 82600
+    },
+    {
+      "epoch": 0.8265,
+      "grad_norm": 50.51581954956055,
+      "learning_rate": 6.69556e-07,
+      "loss": 0.408,
+      "step": 82650
+    },
+    {
+      "epoch": 0.827,
+      "grad_norm": 63.566707611083984,
+      "learning_rate": 6.69356e-07,
+      "loss": 0.4288,
+      "step": 82700
+    },
+    {
+      "epoch": 0.8275,
+      "grad_norm": 3.45732045173645,
+      "learning_rate": 6.691559999999999e-07,
+      "loss": 0.5257,
+      "step": 82750
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 13.433048248291016,
+      "learning_rate": 6.689559999999999e-07,
+      "loss": 0.4753,
+      "step": 82800
+    },
+    {
+      "epoch": 0.8285,
+      "grad_norm": 0.32884418964385986,
+      "learning_rate": 6.68756e-07,
+      "loss": 0.4934,
+      "step": 82850
+    },
+    {
+      "epoch": 0.829,
+      "grad_norm": 13.126819610595703,
+      "learning_rate": 6.68556e-07,
+      "loss": 0.3574,
+      "step": 82900
+    },
+    {
+      "epoch": 0.8295,
+      "grad_norm": 48.53691482543945,
+      "learning_rate": 6.683559999999999e-07,
+      "loss": 0.7363,
+      "step": 82950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 87.21624755859375,
+      "learning_rate": 6.68156e-07,
+      "loss": 0.4372,
+      "step": 83000
+    },
+    {
+      "epoch": 0.8305,
+      "grad_norm": 53.67840576171875,
+      "learning_rate": 6.67956e-07,
+      "loss": 0.5366,
+      "step": 83050
+    },
+    {
+      "epoch": 0.831,
+      "grad_norm": 72.83885192871094,
+      "learning_rate": 6.677559999999999e-07,
+      "loss": 0.5919,
+      "step": 83100
+    },
+    {
+      "epoch": 0.8315,
+      "grad_norm": 24.180238723754883,
+      "learning_rate": 6.67556e-07,
+      "loss": 0.5037,
+      "step": 83150
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 68.72847747802734,
+      "learning_rate": 6.673559999999999e-07,
+      "loss": 0.5177,
+      "step": 83200
+    },
+    {
+      "epoch": 0.8325,
+      "grad_norm": 8.159306526184082,
+      "learning_rate": 6.671559999999999e-07,
+      "loss": 0.4894,
+      "step": 83250
+    },
+    {
+      "epoch": 0.833,
+      "grad_norm": 77.16256713867188,
+      "learning_rate": 6.66956e-07,
+      "loss": 0.4338,
+      "step": 83300
+    },
+    {
+      "epoch": 0.8335,
+      "grad_norm": 15.30505657196045,
+      "learning_rate": 6.66756e-07,
+      "loss": 0.4723,
+      "step": 83350
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 56.12632751464844,
+      "learning_rate": 6.665560000000001e-07,
+      "loss": 0.4587,
+      "step": 83400
+    },
+    {
+      "epoch": 0.8345,
+      "grad_norm": 16.000368118286133,
+      "learning_rate": 6.663559999999999e-07,
+      "loss": 0.5214,
+      "step": 83450
+    },
+    {
+      "epoch": 0.835,
+      "grad_norm": 2.457681655883789,
+      "learning_rate": 6.661559999999999e-07,
+      "loss": 0.5707,
+      "step": 83500
+    },
+    {
+      "epoch": 0.8355,
+      "grad_norm": 107.4878921508789,
+      "learning_rate": 6.65956e-07,
+      "loss": 0.5638,
+      "step": 83550
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 133.30532836914062,
+      "learning_rate": 6.65756e-07,
+      "loss": 0.5926,
+      "step": 83600
+    },
+    {
+      "epoch": 0.8365,
+      "grad_norm": 63.642738342285156,
+      "learning_rate": 6.65556e-07,
+      "loss": 0.3443,
+      "step": 83650
+    },
+    {
+      "epoch": 0.837,
+      "grad_norm": 10.244129180908203,
+      "learning_rate": 6.65356e-07,
+      "loss": 0.4525,
+      "step": 83700
+    },
+    {
+      "epoch": 0.8375,
+      "grad_norm": 21.855087280273438,
+      "learning_rate": 6.65156e-07,
+      "loss": 0.5493,
+      "step": 83750
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 8.698675155639648,
+      "learning_rate": 6.649559999999999e-07,
+      "loss": 0.4376,
+      "step": 83800
+    },
+    {
+      "epoch": 0.8385,
+      "grad_norm": 56.18486785888672,
+      "learning_rate": 6.64756e-07,
+      "loss": 0.4455,
+      "step": 83850
+    },
+    {
+      "epoch": 0.839,
+      "grad_norm": 3.442760705947876,
+      "learning_rate": 6.64556e-07,
+      "loss": 0.3941,
+      "step": 83900
+    },
+    {
+      "epoch": 0.8395,
+      "grad_norm": 108.05015563964844,
+      "learning_rate": 6.643559999999999e-07,
+      "loss": 0.4177,
+      "step": 83950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.338204383850098,
+      "learning_rate": 6.64156e-07,
+      "loss": 0.4218,
+      "step": 84000
+    },
+    {
+      "epoch": 0.8405,
+      "grad_norm": 53.90700149536133,
+      "learning_rate": 6.6396e-07,
+      "loss": 0.3908,
+      "step": 84050
+    },
+    {
+      "epoch": 0.841,
+      "grad_norm": 67.98553466796875,
+      "learning_rate": 6.6376e-07,
+      "loss": 0.3739,
+      "step": 84100
+    },
+    {
+      "epoch": 0.8415,
+      "grad_norm": 45.681941986083984,
+      "learning_rate": 6.6356e-07,
+      "loss": 0.4754,
+      "step": 84150
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 84.54562377929688,
+      "learning_rate": 6.633599999999999e-07,
+      "loss": 0.5086,
+      "step": 84200
+    },
+    {
+      "epoch": 0.8425,
+      "grad_norm": 101.58403778076172,
+      "learning_rate": 6.631599999999999e-07,
+      "loss": 0.4529,
+      "step": 84250
+    },
+    {
+      "epoch": 0.843,
+      "grad_norm": 87.35352325439453,
+      "learning_rate": 6.6296e-07,
+      "loss": 0.561,
+      "step": 84300
+    },
+    {
+      "epoch": 0.8435,
+      "grad_norm": 64.42286682128906,
+      "learning_rate": 6.6276e-07,
+      "loss": 0.4564,
+      "step": 84350
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 12.244241714477539,
+      "learning_rate": 6.6256e-07,
+      "loss": 0.3562,
+      "step": 84400
+    },
+    {
+      "epoch": 0.8445,
+      "grad_norm": 38.84530258178711,
+      "learning_rate": 6.6236e-07,
+      "loss": 0.537,
+      "step": 84450
+    },
+    {
+      "epoch": 0.845,
+      "grad_norm": 71.45441436767578,
+      "learning_rate": 6.621599999999999e-07,
+      "loss": 0.4359,
+      "step": 84500
+    },
+    {
+      "epoch": 0.8455,
+      "grad_norm": 38.31736755371094,
+      "learning_rate": 6.619599999999999e-07,
+      "loss": 0.3036,
+      "step": 84550
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 1.6229268312454224,
+      "learning_rate": 6.6176e-07,
+      "loss": 0.5366,
+      "step": 84600
+    },
+    {
+      "epoch": 0.8465,
+      "grad_norm": 65.19337463378906,
+      "learning_rate": 6.6156e-07,
+      "loss": 0.5766,
+      "step": 84650
+    },
+    {
+      "epoch": 0.847,
+      "grad_norm": 96.37982940673828,
+      "learning_rate": 6.613599999999999e-07,
+      "loss": 0.4957,
+      "step": 84700
+    },
+    {
+      "epoch": 0.8475,
+      "grad_norm": 117.63880920410156,
+      "learning_rate": 6.6116e-07,
+      "loss": 0.5433,
+      "step": 84750
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 107.3805923461914,
+      "learning_rate": 6.6096e-07,
+      "loss": 0.4758,
+      "step": 84800
+    },
+    {
+      "epoch": 0.8485,
+      "grad_norm": 105.60696411132812,
+      "learning_rate": 6.607599999999999e-07,
+      "loss": 0.562,
+      "step": 84850
+    },
+    {
+      "epoch": 0.849,
+      "grad_norm": 0.5521818995475769,
+      "learning_rate": 6.6056e-07,
+      "loss": 0.3867,
+      "step": 84900
+    },
+    {
+      "epoch": 0.8495,
+      "grad_norm": 15.534207344055176,
+      "learning_rate": 6.603599999999999e-07,
+      "loss": 0.3862,
+      "step": 84950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.9523701667785645,
+      "learning_rate": 6.601599999999999e-07,
+      "loss": 0.4487,
+      "step": 85000
+    },
+    {
+      "epoch": 0.8505,
+      "grad_norm": 107.40753936767578,
+      "learning_rate": 6.5996e-07,
+      "loss": 0.4253,
+      "step": 85050
+    },
+    {
+      "epoch": 0.851,
+      "grad_norm": 2.598055362701416,
+      "learning_rate": 6.5976e-07,
+      "loss": 0.5241,
+      "step": 85100
+    },
+    {
+      "epoch": 0.8515,
+      "grad_norm": 55.7864875793457,
+      "learning_rate": 6.5956e-07,
+      "loss": 0.5934,
+      "step": 85150
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 4.6609697341918945,
+      "learning_rate": 6.593599999999999e-07,
+      "loss": 0.4759,
+      "step": 85200
+    },
+    {
+      "epoch": 0.8525,
+      "grad_norm": 103.4588851928711,
+      "learning_rate": 6.591599999999999e-07,
+      "loss": 0.4338,
+      "step": 85250
+    },
+    {
+      "epoch": 0.853,
+      "grad_norm": 4.739608287811279,
+      "learning_rate": 6.589599999999999e-07,
+      "loss": 0.5771,
+      "step": 85300
+    },
+    {
+      "epoch": 0.8535,
+      "grad_norm": 81.03630828857422,
+      "learning_rate": 6.5876e-07,
+      "loss": 0.4308,
+      "step": 85350
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 10.07264232635498,
+      "learning_rate": 6.5856e-07,
+      "loss": 0.5954,
+      "step": 85400
+    },
+    {
+      "epoch": 0.8545,
+      "grad_norm": 87.12622833251953,
+      "learning_rate": 6.583599999999999e-07,
+      "loss": 0.5724,
+      "step": 85450
+    },
+    {
+      "epoch": 0.855,
+      "grad_norm": 12.88304615020752,
+      "learning_rate": 6.5816e-07,
+      "loss": 0.3812,
+      "step": 85500
+    },
+    {
+      "epoch": 0.8555,
+      "grad_norm": 2.8463549613952637,
+      "learning_rate": 6.579599999999999e-07,
+      "loss": 0.5091,
+      "step": 85550
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 107.44549560546875,
+      "learning_rate": 6.577599999999999e-07,
+      "loss": 0.5983,
+      "step": 85600
+    },
+    {
+      "epoch": 0.8565,
+      "grad_norm": 25.965686798095703,
+      "learning_rate": 6.5756e-07,
+      "loss": 0.4518,
+      "step": 85650
+    },
+    {
+      "epoch": 0.857,
+      "grad_norm": 13.483121871948242,
+      "learning_rate": 6.5736e-07,
+      "loss": 0.4494,
+      "step": 85700
+    },
+    {
+      "epoch": 0.8575,
+      "grad_norm": 57.970027923583984,
+      "learning_rate": 6.571599999999999e-07,
+      "loss": 0.5211,
+      "step": 85750
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 61.619537353515625,
+      "learning_rate": 6.5696e-07,
+      "loss": 0.4131,
+      "step": 85800
+    },
+    {
+      "epoch": 0.8585,
+      "grad_norm": 72.63780975341797,
+      "learning_rate": 6.5676e-07,
+      "loss": 0.4786,
+      "step": 85850
+    },
+    {
+      "epoch": 0.859,
+      "grad_norm": 32.84150314331055,
+      "learning_rate": 6.5656e-07,
+      "loss": 0.3868,
+      "step": 85900
+    },
+    {
+      "epoch": 0.8595,
+      "grad_norm": 11.908856391906738,
+      "learning_rate": 6.5636e-07,
+      "loss": 0.4036,
+      "step": 85950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 124.65275573730469,
+      "learning_rate": 6.561599999999999e-07,
+      "loss": 0.5248,
+      "step": 86000
+    },
+    {
+      "epoch": 0.8605,
+      "grad_norm": 77.87248992919922,
+      "learning_rate": 6.5596e-07,
+      "loss": 0.631,
+      "step": 86050
+    },
+    {
+      "epoch": 0.861,
+      "grad_norm": 88.47559356689453,
+      "learning_rate": 6.5576e-07,
+      "loss": 0.5183,
+      "step": 86100
+    },
+    {
+      "epoch": 0.8615,
+      "grad_norm": 35.03434371948242,
+      "learning_rate": 6.5556e-07,
+      "loss": 0.536,
+      "step": 86150
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 49.06086349487305,
+      "learning_rate": 6.553600000000001e-07,
+      "loss": 0.4981,
+      "step": 86200
+    },
+    {
+      "epoch": 0.8625,
+      "grad_norm": 22.367053985595703,
+      "learning_rate": 6.551599999999999e-07,
+      "loss": 0.5554,
+      "step": 86250
+    },
+    {
+      "epoch": 0.863,
+      "grad_norm": 46.18075180053711,
+      "learning_rate": 6.549599999999999e-07,
+      "loss": 0.4415,
+      "step": 86300
+    },
+    {
+      "epoch": 0.8635,
+      "grad_norm": 85.1124496459961,
+      "learning_rate": 6.5476e-07,
+      "loss": 0.6429,
+      "step": 86350
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 75.1630630493164,
+      "learning_rate": 6.5456e-07,
+      "loss": 0.4281,
+      "step": 86400
+    },
+    {
+      "epoch": 0.8645,
+      "grad_norm": 23.679262161254883,
+      "learning_rate": 6.5436e-07,
+      "loss": 0.4745,
+      "step": 86450
+    },
+    {
+      "epoch": 0.865,
+      "grad_norm": 40.03208541870117,
+      "learning_rate": 6.5416e-07,
+      "loss": 0.3905,
+      "step": 86500
+    },
+    {
+      "epoch": 0.8655,
+      "grad_norm": 65.0140151977539,
+      "learning_rate": 6.5396e-07,
+      "loss": 0.6504,
+      "step": 86550
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 85.9790267944336,
+      "learning_rate": 6.537599999999999e-07,
+      "loss": 0.3979,
+      "step": 86600
+    },
+    {
+      "epoch": 0.8665,
+      "grad_norm": 3.0157697200775146,
+      "learning_rate": 6.5356e-07,
+      "loss": 0.4499,
+      "step": 86650
+    },
+    {
+      "epoch": 0.867,
+      "grad_norm": 86.25167846679688,
+      "learning_rate": 6.5336e-07,
+      "loss": 0.5585,
+      "step": 86700
+    },
+    {
+      "epoch": 0.8675,
+      "grad_norm": 26.297157287597656,
+      "learning_rate": 6.531599999999999e-07,
+      "loss": 0.5065,
+      "step": 86750
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 53.790348052978516,
+      "learning_rate": 6.5296e-07,
+      "loss": 0.507,
+      "step": 86800
+    },
+    {
+      "epoch": 0.8685,
+      "grad_norm": 45.35681915283203,
+      "learning_rate": 6.5276e-07,
+      "loss": 0.4223,
+      "step": 86850
+    },
+    {
+      "epoch": 0.869,
+      "grad_norm": 108.29571533203125,
+      "learning_rate": 6.5256e-07,
+      "loss": 0.4757,
+      "step": 86900
+    },
+    {
+      "epoch": 0.8695,
+      "grad_norm": 1.806410312652588,
+      "learning_rate": 6.5236e-07,
+      "loss": 0.4832,
+      "step": 86950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 46.172607421875,
+      "learning_rate": 6.521599999999999e-07,
+      "loss": 0.599,
+      "step": 87000
+    },
+    {
+      "epoch": 0.8705,
+      "grad_norm": 37.954742431640625,
+      "learning_rate": 6.519599999999999e-07,
+      "loss": 0.6213,
+      "step": 87050
+    },
+    {
+      "epoch": 0.871,
+      "grad_norm": 11.543988227844238,
+      "learning_rate": 6.5176e-07,
+      "loss": 0.4723,
+      "step": 87100
+    },
+    {
+      "epoch": 0.8715,
+      "grad_norm": 26.00443458557129,
+      "learning_rate": 6.5156e-07,
+      "loss": 0.4994,
+      "step": 87150
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 144.9601593017578,
+      "learning_rate": 6.5136e-07,
+      "loss": 0.5887,
+      "step": 87200
+    },
+    {
+      "epoch": 0.8725,
+      "grad_norm": 53.76115417480469,
+      "learning_rate": 6.5116e-07,
+      "loss": 0.4368,
+      "step": 87250
+    },
+    {
+      "epoch": 0.873,
+      "grad_norm": 56.003292083740234,
+      "learning_rate": 6.5096e-07,
+      "loss": 0.49,
+      "step": 87300
+    },
+    {
+      "epoch": 0.8735,
+      "grad_norm": 81.86634063720703,
+      "learning_rate": 6.507599999999999e-07,
+      "loss": 0.4194,
+      "step": 87350
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 46.207969665527344,
+      "learning_rate": 6.5056e-07,
+      "loss": 0.5626,
+      "step": 87400
+    },
+    {
+      "epoch": 0.8745,
+      "grad_norm": 93.14911651611328,
+      "learning_rate": 6.5036e-07,
+      "loss": 0.5803,
+      "step": 87450
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 91.9137954711914,
+      "learning_rate": 6.501599999999999e-07,
+      "loss": 0.4872,
+      "step": 87500
+    },
+    {
+      "epoch": 0.8755,
+      "grad_norm": 73.91366577148438,
+      "learning_rate": 6.4996e-07,
+      "loss": 0.5283,
+      "step": 87550
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 35.279666900634766,
+      "learning_rate": 6.4976e-07,
+      "loss": 0.4378,
+      "step": 87600
+    },
+    {
+      "epoch": 0.8765,
+      "grad_norm": 15.952431678771973,
+      "learning_rate": 6.4956e-07,
+      "loss": 0.4564,
+      "step": 87650
+    },
+    {
+      "epoch": 0.877,
+      "grad_norm": 2.8610007762908936,
+      "learning_rate": 6.4936e-07,
+      "loss": 0.4478,
+      "step": 87700
+    },
+    {
+      "epoch": 0.8775,
+      "grad_norm": 64.05665588378906,
+      "learning_rate": 6.491599999999999e-07,
+      "loss": 0.4061,
+      "step": 87750
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 3.5503172874450684,
+      "learning_rate": 6.489599999999999e-07,
+      "loss": 0.4837,
+      "step": 87800
+    },
+    {
+      "epoch": 0.8785,
+      "grad_norm": 11.314192771911621,
+      "learning_rate": 6.4876e-07,
+      "loss": 0.3553,
+      "step": 87850
+    },
+    {
+      "epoch": 0.879,
+      "grad_norm": 23.217226028442383,
+      "learning_rate": 6.4856e-07,
+      "loss": 0.4392,
+      "step": 87900
+    },
+    {
+      "epoch": 0.8795,
+      "grad_norm": 2.8893821239471436,
+      "learning_rate": 6.483600000000001e-07,
+      "loss": 0.4047,
+      "step": 87950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.448762893676758,
+      "learning_rate": 6.4816e-07,
+      "loss": 0.4604,
+      "step": 88000
+    },
+    {
+      "epoch": 0.8805,
+      "grad_norm": 84.17955017089844,
+      "learning_rate": 6.479599999999999e-07,
+      "loss": 0.4875,
+      "step": 88050
+    },
+    {
+      "epoch": 0.881,
+      "grad_norm": 68.11402130126953,
+      "learning_rate": 6.4776e-07,
+      "loss": 0.5875,
+      "step": 88100
+    },
+    {
+      "epoch": 0.8815,
+      "grad_norm": 77.2765884399414,
+      "learning_rate": 6.4756e-07,
+      "loss": 0.4419,
+      "step": 88150
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 16.410972595214844,
+      "learning_rate": 6.4736e-07,
+      "loss": 0.4844,
+      "step": 88200
+    },
+    {
+      "epoch": 0.8825,
+      "grad_norm": 31.93629264831543,
+      "learning_rate": 6.4716e-07,
+      "loss": 0.4829,
+      "step": 88250
+    },
+    {
+      "epoch": 0.883,
+      "grad_norm": 68.09761047363281,
+      "learning_rate": 6.4696e-07,
+      "loss": 0.3913,
+      "step": 88300
+    },
+    {
+      "epoch": 0.8835,
+      "grad_norm": 91.0275650024414,
+      "learning_rate": 6.4676e-07,
+      "loss": 0.6128,
+      "step": 88350
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 67.89716339111328,
+      "learning_rate": 6.4656e-07,
+      "loss": 0.5768,
+      "step": 88400
+    },
+    {
+      "epoch": 0.8845,
+      "grad_norm": 21.36505126953125,
+      "learning_rate": 6.4636e-07,
+      "loss": 0.3613,
+      "step": 88450
+    },
+    {
+      "epoch": 0.885,
+      "grad_norm": 2.784221887588501,
+      "learning_rate": 6.461599999999999e-07,
+      "loss": 0.3029,
+      "step": 88500
+    },
+    {
+      "epoch": 0.8855,
+      "grad_norm": 4.084001064300537,
+      "learning_rate": 6.4596e-07,
+      "loss": 0.3698,
+      "step": 88550
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 72.70446014404297,
+      "learning_rate": 6.4576e-07,
+      "loss": 0.4915,
+      "step": 88600
+    },
+    {
+      "epoch": 0.8865,
+      "grad_norm": 55.41581726074219,
+      "learning_rate": 6.4556e-07,
+      "loss": 0.5785,
+      "step": 88650
+    },
+    {
+      "epoch": 0.887,
+      "grad_norm": 35.189048767089844,
+      "learning_rate": 6.453600000000001e-07,
+      "loss": 0.5671,
+      "step": 88700
+    },
+    {
+      "epoch": 0.8875,
+      "grad_norm": 57.02684783935547,
+      "learning_rate": 6.451599999999999e-07,
+      "loss": 0.4981,
+      "step": 88750
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 0.43074750900268555,
+      "learning_rate": 6.449599999999999e-07,
+      "loss": 0.4555,
+      "step": 88800
+    },
+    {
+      "epoch": 0.8885,
+      "grad_norm": 13.582013130187988,
+      "learning_rate": 6.4476e-07,
+      "loss": 0.3789,
+      "step": 88850
+    },
+    {
+      "epoch": 0.889,
+      "grad_norm": 65.70917510986328,
+      "learning_rate": 6.4456e-07,
+      "loss": 0.5623,
+      "step": 88900
+    },
+    {
+      "epoch": 0.8895,
+      "grad_norm": 48.792179107666016,
+      "learning_rate": 6.4436e-07,
+      "loss": 0.5911,
+      "step": 88950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.270796775817871,
+      "learning_rate": 6.4416e-07,
+      "loss": 0.4401,
+      "step": 89000
+    },
+    {
+      "epoch": 0.8905,
+      "grad_norm": 89.74335479736328,
+      "learning_rate": 6.4396e-07,
+      "loss": 0.3221,
+      "step": 89050
+    },
+    {
+      "epoch": 0.891,
+      "grad_norm": 46.43724822998047,
+      "learning_rate": 6.437599999999999e-07,
+      "loss": 0.3968,
+      "step": 89100
+    },
+    {
+      "epoch": 0.8915,
+      "grad_norm": 2.0606515407562256,
+      "learning_rate": 6.4356e-07,
+      "loss": 0.447,
+      "step": 89150
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 91.6527099609375,
+      "learning_rate": 6.4336e-07,
+      "loss": 0.4425,
+      "step": 89200
+    },
+    {
+      "epoch": 0.8925,
+      "grad_norm": 42.8084716796875,
+      "learning_rate": 6.431599999999999e-07,
+      "loss": 0.4977,
+      "step": 89250
+    },
+    {
+      "epoch": 0.893,
+      "grad_norm": 13.197081565856934,
+      "learning_rate": 6.4296e-07,
+      "loss": 0.395,
+      "step": 89300
+    },
+    {
+      "epoch": 0.8935,
+      "grad_norm": 97.76970672607422,
+      "learning_rate": 6.4276e-07,
+      "loss": 0.4206,
+      "step": 89350
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 103.70655822753906,
+      "learning_rate": 6.4256e-07,
+      "loss": 0.6174,
+      "step": 89400
+    },
+    {
+      "epoch": 0.8945,
+      "grad_norm": 25.918859481811523,
+      "learning_rate": 6.4236e-07,
+      "loss": 0.5479,
+      "step": 89450
+    },
+    {
+      "epoch": 0.895,
+      "grad_norm": 28.53264808654785,
+      "learning_rate": 6.421599999999999e-07,
+      "loss": 0.5709,
+      "step": 89500
+    },
+    {
+      "epoch": 0.8955,
+      "grad_norm": 38.950416564941406,
+      "learning_rate": 6.419599999999999e-07,
+      "loss": 0.5301,
+      "step": 89550
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 95.71314239501953,
+      "learning_rate": 6.4176e-07,
+      "loss": 0.5001,
+      "step": 89600
+    },
+    {
+      "epoch": 0.8965,
+      "grad_norm": 7.237802505493164,
+      "learning_rate": 6.4156e-07,
+      "loss": 0.5101,
+      "step": 89650
+    },
+    {
+      "epoch": 0.897,
+      "grad_norm": 104.18899536132812,
+      "learning_rate": 6.4136e-07,
+      "loss": 0.445,
+      "step": 89700
+    },
+    {
+      "epoch": 0.8975,
+      "grad_norm": 88.33003997802734,
+      "learning_rate": 6.4116e-07,
+      "loss": 0.4632,
+      "step": 89750
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 0.11056256294250488,
+      "learning_rate": 6.409599999999999e-07,
+      "loss": 0.4456,
+      "step": 89800
+    },
+    {
+      "epoch": 0.8985,
+      "grad_norm": 23.158008575439453,
+      "learning_rate": 6.407599999999999e-07,
+      "loss": 0.5806,
+      "step": 89850
+    },
+    {
+      "epoch": 0.899,
+      "grad_norm": 45.30116271972656,
+      "learning_rate": 6.4056e-07,
+      "loss": 0.4927,
+      "step": 89900
+    },
+    {
+      "epoch": 0.8995,
+      "grad_norm": 13.431071281433105,
+      "learning_rate": 6.4036e-07,
+      "loss": 0.5075,
+      "step": 89950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 42.31669235229492,
+      "learning_rate": 6.401599999999999e-07,
+      "loss": 0.4129,
+      "step": 90000
+    },
+    {
+      "epoch": 0.9005,
+      "grad_norm": 35.61188888549805,
+      "learning_rate": 6.3996e-07,
+      "loss": 0.5002,
+      "step": 90050
+    },
+    {
+      "epoch": 0.901,
+      "grad_norm": 33.90731430053711,
+      "learning_rate": 6.3976e-07,
+      "loss": 0.4587,
+      "step": 90100
+    },
+    {
+      "epoch": 0.9015,
+      "grad_norm": 17.363262176513672,
+      "learning_rate": 6.395599999999999e-07,
+      "loss": 0.3567,
+      "step": 90150
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 88.20648193359375,
+      "learning_rate": 6.3936e-07,
+      "loss": 0.3825,
+      "step": 90200
+    },
+    {
+      "epoch": 0.9025,
+      "grad_norm": 105.39450073242188,
+      "learning_rate": 6.391599999999999e-07,
+      "loss": 0.4482,
+      "step": 90250
+    },
+    {
+      "epoch": 0.903,
+      "grad_norm": 41.72043991088867,
+      "learning_rate": 6.389599999999999e-07,
+      "loss": 0.4216,
+      "step": 90300
+    },
+    {
+      "epoch": 0.9035,
+      "grad_norm": 52.67734909057617,
+      "learning_rate": 6.3876e-07,
+      "loss": 0.5691,
+      "step": 90350
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 2.5297904014587402,
+      "learning_rate": 6.3856e-07,
+      "loss": 0.4965,
+      "step": 90400
+    },
+    {
+      "epoch": 0.9045,
+      "grad_norm": 100.61029815673828,
+      "learning_rate": 6.383600000000001e-07,
+      "loss": 0.3907,
+      "step": 90450
+    },
+    {
+      "epoch": 0.905,
+      "grad_norm": 123.27056884765625,
+      "learning_rate": 6.381599999999999e-07,
+      "loss": 0.6531,
+      "step": 90500
+    },
+    {
+      "epoch": 0.9055,
+      "grad_norm": 57.43628692626953,
+      "learning_rate": 6.379599999999999e-07,
+      "loss": 0.5612,
+      "step": 90550
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 6.510050296783447,
+      "learning_rate": 6.3776e-07,
+      "loss": 0.6138,
+      "step": 90600
+    },
+    {
+      "epoch": 0.9065,
+      "grad_norm": 71.36262512207031,
+      "learning_rate": 6.3756e-07,
+      "loss": 0.3697,
+      "step": 90650
+    },
+    {
+      "epoch": 0.907,
+      "grad_norm": 92.18431091308594,
+      "learning_rate": 6.3736e-07,
+      "loss": 0.4635,
+      "step": 90700
+    },
+    {
+      "epoch": 0.9075,
+      "grad_norm": 73.8544692993164,
+      "learning_rate": 6.3716e-07,
+      "loss": 0.4039,
+      "step": 90750
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 84.32909393310547,
+      "learning_rate": 6.3696e-07,
+      "loss": 0.5271,
+      "step": 90800
+    },
+    {
+      "epoch": 0.9085,
+      "grad_norm": 71.52722930908203,
+      "learning_rate": 6.367599999999999e-07,
+      "loss": 0.4169,
+      "step": 90850
+    },
+    {
+      "epoch": 0.909,
+      "grad_norm": 10.915364265441895,
+      "learning_rate": 6.36564e-07,
+      "loss": 0.467,
+      "step": 90900
+    },
+    {
+      "epoch": 0.9095,
+      "grad_norm": 74.15825653076172,
+      "learning_rate": 6.36364e-07,
+      "loss": 0.4426,
+      "step": 90950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 91.29158020019531,
+      "learning_rate": 6.361639999999999e-07,
+      "loss": 0.6179,
+      "step": 91000
+    },
+    {
+      "epoch": 0.9105,
+      "grad_norm": 0.4594174027442932,
+      "learning_rate": 6.35964e-07,
+      "loss": 0.5343,
+      "step": 91050
+    },
+    {
+      "epoch": 0.911,
+      "grad_norm": 124.07490539550781,
+      "learning_rate": 6.35764e-07,
+      "loss": 0.4171,
+      "step": 91100
+    },
+    {
+      "epoch": 0.9115,
+      "grad_norm": 43.403846740722656,
+      "learning_rate": 6.35564e-07,
+      "loss": 0.5054,
+      "step": 91150
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 43.1138801574707,
+      "learning_rate": 6.35364e-07,
+      "loss": 0.4225,
+      "step": 91200
+    },
+    {
+      "epoch": 0.9125,
+      "grad_norm": 117.5290756225586,
+      "learning_rate": 6.351639999999999e-07,
+      "loss": 0.5131,
+      "step": 91250
+    },
+    {
+      "epoch": 0.913,
+      "grad_norm": 67.10313415527344,
+      "learning_rate": 6.349679999999999e-07,
+      "loss": 0.3869,
+      "step": 91300
+    },
+    {
+      "epoch": 0.9135,
+      "grad_norm": 12.4064359664917,
+      "learning_rate": 6.34768e-07,
+      "loss": 0.4939,
+      "step": 91350
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 20.02683448791504,
+      "learning_rate": 6.34568e-07,
+      "loss": 0.5706,
+      "step": 91400
+    },
+    {
+      "epoch": 0.9145,
+      "grad_norm": 64.59272766113281,
+      "learning_rate": 6.34368e-07,
+      "loss": 0.3377,
+      "step": 91450
+    },
+    {
+      "epoch": 0.915,
+      "grad_norm": 2.4483537673950195,
+      "learning_rate": 6.34168e-07,
+      "loss": 0.349,
+      "step": 91500
+    },
+    {
+      "epoch": 0.9155,
+      "grad_norm": 0.7086917757987976,
+      "learning_rate": 6.33968e-07,
+      "loss": 0.4049,
+      "step": 91550
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 17.961669921875,
+      "learning_rate": 6.337679999999999e-07,
+      "loss": 0.3951,
+      "step": 91600
+    },
+    {
+      "epoch": 0.9165,
+      "grad_norm": 62.844703674316406,
+      "learning_rate": 6.33568e-07,
+      "loss": 0.4574,
+      "step": 91650
+    },
+    {
+      "epoch": 0.917,
+      "grad_norm": 4.404021263122559,
+      "learning_rate": 6.33372e-07,
+      "loss": 0.4018,
+      "step": 91700
+    },
+    {
+      "epoch": 0.9175,
+      "grad_norm": 1.865087866783142,
+      "learning_rate": 6.331719999999999e-07,
+      "loss": 0.5116,
+      "step": 91750
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 49.26223373413086,
+      "learning_rate": 6.32976e-07,
+      "loss": 0.5372,
+      "step": 91800
+    },
+    {
+      "epoch": 0.9185,
+      "grad_norm": 87.36351013183594,
+      "learning_rate": 6.32776e-07,
+      "loss": 0.594,
+      "step": 91850
+    },
+    {
+      "epoch": 0.919,
+      "grad_norm": 13.864863395690918,
+      "learning_rate": 6.32576e-07,
+      "loss": 0.385,
+      "step": 91900
+    },
+    {
+      "epoch": 0.9195,
+      "grad_norm": 85.708251953125,
+      "learning_rate": 6.32376e-07,
+      "loss": 0.4528,
+      "step": 91950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 66.8047866821289,
+      "learning_rate": 6.321759999999999e-07,
+      "loss": 0.5978,
+      "step": 92000
+    },
+    {
+      "epoch": 0.9205,
+      "grad_norm": 91.00402069091797,
+      "learning_rate": 6.319759999999999e-07,
+      "loss": 0.512,
+      "step": 92050
+    },
+    {
+      "epoch": 0.921,
+      "grad_norm": 56.01173400878906,
+      "learning_rate": 6.31776e-07,
+      "loss": 0.5682,
+      "step": 92100
+    },
+    {
+      "epoch": 0.9215,
+      "grad_norm": 71.80009460449219,
+      "learning_rate": 6.31576e-07,
+      "loss": 0.5471,
+      "step": 92150
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 49.644187927246094,
+      "learning_rate": 6.31376e-07,
+      "loss": 0.6007,
+      "step": 92200
+    },
+    {
+      "epoch": 0.9225,
+      "grad_norm": 6.2205119132995605,
+      "learning_rate": 6.31176e-07,
+      "loss": 0.3881,
+      "step": 92250
+    },
+    {
+      "epoch": 0.923,
+      "grad_norm": 34.30057907104492,
+      "learning_rate": 6.309759999999999e-07,
+      "loss": 0.4156,
+      "step": 92300
+    },
+    {
+      "epoch": 0.9235,
+      "grad_norm": 65.61951446533203,
+      "learning_rate": 6.307759999999999e-07,
+      "loss": 0.5385,
+      "step": 92350
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 96.61206817626953,
+      "learning_rate": 6.30576e-07,
+      "loss": 0.5514,
+      "step": 92400
+    },
+    {
+      "epoch": 0.9245,
+      "grad_norm": 21.83513641357422,
+      "learning_rate": 6.30376e-07,
+      "loss": 0.5608,
+      "step": 92450
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 17.049848556518555,
+      "learning_rate": 6.301759999999999e-07,
+      "loss": 0.4524,
+      "step": 92500
+    },
+    {
+      "epoch": 0.9255,
+      "grad_norm": 19.969083786010742,
+      "learning_rate": 6.29976e-07,
+      "loss": 0.5688,
+      "step": 92550
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 52.34061050415039,
+      "learning_rate": 6.29776e-07,
+      "loss": 0.476,
+      "step": 92600
+    },
+    {
+      "epoch": 0.9265,
+      "grad_norm": 86.44994354248047,
+      "learning_rate": 6.295759999999999e-07,
+      "loss": 0.3973,
+      "step": 92650
+    },
+    {
+      "epoch": 0.927,
+      "grad_norm": 19.09813117980957,
+      "learning_rate": 6.29376e-07,
+      "loss": 0.4407,
+      "step": 92700
+    },
+    {
+      "epoch": 0.9275,
+      "grad_norm": 17.017074584960938,
+      "learning_rate": 6.291759999999999e-07,
+      "loss": 0.4964,
+      "step": 92750
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 68.21665954589844,
+      "learning_rate": 6.289759999999999e-07,
+      "loss": 0.4977,
+      "step": 92800
+    },
+    {
+      "epoch": 0.9285,
+      "grad_norm": 15.754688262939453,
+      "learning_rate": 6.28776e-07,
+      "loss": 0.5211,
+      "step": 92850
+    },
+    {
+      "epoch": 0.929,
+      "grad_norm": 39.87798309326172,
+      "learning_rate": 6.28576e-07,
+      "loss": 0.4764,
+      "step": 92900
+    },
+    {
+      "epoch": 0.9295,
+      "grad_norm": 66.07242584228516,
+      "learning_rate": 6.283760000000001e-07,
+      "loss": 0.4276,
+      "step": 92950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 68.16017150878906,
+      "learning_rate": 6.281759999999999e-07,
+      "loss": 0.4338,
+      "step": 93000
+    },
+    {
+      "epoch": 0.9305,
+      "grad_norm": 92.84886932373047,
+      "learning_rate": 6.279759999999999e-07,
+      "loss": 0.3994,
+      "step": 93050
+    },
+    {
+      "epoch": 0.931,
+      "grad_norm": 7.953521251678467,
+      "learning_rate": 6.27776e-07,
+      "loss": 0.5969,
+      "step": 93100
+    },
+    {
+      "epoch": 0.9315,
+      "grad_norm": 6.302267551422119,
+      "learning_rate": 6.27576e-07,
+      "loss": 0.5746,
+      "step": 93150
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 57.08115005493164,
+      "learning_rate": 6.27376e-07,
+      "loss": 0.6096,
+      "step": 93200
+    },
+    {
+      "epoch": 0.9325,
+      "grad_norm": 74.17098236083984,
+      "learning_rate": 6.27176e-07,
+      "loss": 0.5127,
+      "step": 93250
+    },
+    {
+      "epoch": 0.933,
+      "grad_norm": 3.4939818382263184,
+      "learning_rate": 6.26976e-07,
+      "loss": 0.4822,
+      "step": 93300
+    },
+    {
+      "epoch": 0.9335,
+      "grad_norm": 15.249403953552246,
+      "learning_rate": 6.26776e-07,
+      "loss": 0.4677,
+      "step": 93350
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 8.158519744873047,
+      "learning_rate": 6.26576e-07,
+      "loss": 0.637,
+      "step": 93400
+    },
+    {
+      "epoch": 0.9345,
+      "grad_norm": 3.797173023223877,
+      "learning_rate": 6.26376e-07,
+      "loss": 0.6001,
+      "step": 93450
+    },
+    {
+      "epoch": 0.935,
+      "grad_norm": 9.297698020935059,
+      "learning_rate": 6.261759999999999e-07,
+      "loss": 0.4783,
+      "step": 93500
+    },
+    {
+      "epoch": 0.9355,
+      "grad_norm": 97.72147369384766,
+      "learning_rate": 6.25976e-07,
+      "loss": 0.5935,
+      "step": 93550
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 30.139896392822266,
+      "learning_rate": 6.25776e-07,
+      "loss": 0.4528,
+      "step": 93600
+    },
+    {
+      "epoch": 0.9365,
+      "grad_norm": 93.17426300048828,
+      "learning_rate": 6.25576e-07,
+      "loss": 0.519,
+      "step": 93650
+    },
+    {
+      "epoch": 0.937,
+      "grad_norm": 105.38163757324219,
+      "learning_rate": 6.253760000000001e-07,
+      "loss": 0.5527,
+      "step": 93700
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 68.39093780517578,
+      "learning_rate": 6.251759999999999e-07,
+      "loss": 0.4141,
+      "step": 93750
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 70.97985076904297,
+      "learning_rate": 6.249759999999999e-07,
+      "loss": 0.4535,
+      "step": 93800
+    },
+    {
+      "epoch": 0.9385,
+      "grad_norm": 74.9505615234375,
+      "learning_rate": 6.24776e-07,
+      "loss": 0.5361,
+      "step": 93850
+    },
+    {
+      "epoch": 0.939,
+      "grad_norm": 21.2496395111084,
+      "learning_rate": 6.24576e-07,
+      "loss": 0.4383,
+      "step": 93900
+    },
+    {
+      "epoch": 0.9395,
+      "grad_norm": 69.52516174316406,
+      "learning_rate": 6.24376e-07,
+      "loss": 0.3776,
+      "step": 93950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 42.33454132080078,
+      "learning_rate": 6.24176e-07,
+      "loss": 0.4317,
+      "step": 94000
+    },
+    {
+      "epoch": 0.9405,
+      "grad_norm": 3.589440107345581,
+      "learning_rate": 6.23976e-07,
+      "loss": 0.4107,
+      "step": 94050
+    },
+    {
+      "epoch": 0.941,
+      "grad_norm": 108.18557739257812,
+      "learning_rate": 6.237759999999999e-07,
+      "loss": 0.4974,
+      "step": 94100
+    },
+    {
+      "epoch": 0.9415,
+      "grad_norm": 17.219356536865234,
+      "learning_rate": 6.23576e-07,
+      "loss": 0.4116,
+      "step": 94150
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 19.18585777282715,
+      "learning_rate": 6.23376e-07,
+      "loss": 0.4121,
+      "step": 94200
+    },
+    {
+      "epoch": 0.9425,
+      "grad_norm": 118.1676254272461,
+      "learning_rate": 6.231759999999999e-07,
+      "loss": 0.5545,
+      "step": 94250
+    },
+    {
+      "epoch": 0.943,
+      "grad_norm": 47.58047103881836,
+      "learning_rate": 6.22976e-07,
+      "loss": 0.4113,
+      "step": 94300
+    },
+    {
+      "epoch": 0.9435,
+      "grad_norm": 81.94721221923828,
+      "learning_rate": 6.22776e-07,
+      "loss": 0.5893,
+      "step": 94350
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 4.948315143585205,
+      "learning_rate": 6.22576e-07,
+      "loss": 0.3997,
+      "step": 94400
+    },
+    {
+      "epoch": 0.9445,
+      "grad_norm": 14.120973587036133,
+      "learning_rate": 6.22376e-07,
+      "loss": 0.4992,
+      "step": 94450
+    },
+    {
+      "epoch": 0.945,
+      "grad_norm": 63.99724197387695,
+      "learning_rate": 6.221759999999999e-07,
+      "loss": 0.5603,
+      "step": 94500
+    },
+    {
+      "epoch": 0.9455,
+      "grad_norm": 46.79153060913086,
+      "learning_rate": 6.219759999999999e-07,
+      "loss": 0.4607,
+      "step": 94550
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 8.22097110748291,
+      "learning_rate": 6.21776e-07,
+      "loss": 0.557,
+      "step": 94600
+    },
+    {
+      "epoch": 0.9465,
+      "grad_norm": 98.2745361328125,
+      "learning_rate": 6.21576e-07,
+      "loss": 0.5769,
+      "step": 94650
+    },
+    {
+      "epoch": 0.947,
+      "grad_norm": 40.03475570678711,
+      "learning_rate": 6.21376e-07,
+      "loss": 0.5056,
+      "step": 94700
+    },
+    {
+      "epoch": 0.9475,
+      "grad_norm": 74.36143493652344,
+      "learning_rate": 6.21176e-07,
+      "loss": 0.4523,
+      "step": 94750
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 51.59547805786133,
+      "learning_rate": 6.209759999999999e-07,
+      "loss": 0.5666,
+      "step": 94800
+    },
+    {
+      "epoch": 0.9485,
+      "grad_norm": 61.492488861083984,
+      "learning_rate": 6.207759999999999e-07,
+      "loss": 0.536,
+      "step": 94850
+    },
+    {
+      "epoch": 0.949,
+      "grad_norm": 3.265573024749756,
+      "learning_rate": 6.20576e-07,
+      "loss": 0.3787,
+      "step": 94900
+    },
+    {
+      "epoch": 0.9495,
+      "grad_norm": 29.715877532958984,
+      "learning_rate": 6.20376e-07,
+      "loss": 0.5144,
+      "step": 94950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.610234260559082,
+      "learning_rate": 6.201759999999999e-07,
+      "loss": 0.4496,
+      "step": 95000
+    },
+    {
+      "epoch": 0.9505,
+      "grad_norm": 3.437190532684326,
+      "learning_rate": 6.19976e-07,
+      "loss": 0.5684,
+      "step": 95050
+    },
+    {
+      "epoch": 0.951,
+      "grad_norm": 48.969093322753906,
+      "learning_rate": 6.19776e-07,
+      "loss": 0.4846,
+      "step": 95100
+    },
+    {
+      "epoch": 0.9515,
+      "grad_norm": 72.4744873046875,
+      "learning_rate": 6.195759999999999e-07,
+      "loss": 0.4036,
+      "step": 95150
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 2.2708630561828613,
+      "learning_rate": 6.19376e-07,
+      "loss": 0.4133,
+      "step": 95200
+    },
+    {
+      "epoch": 0.9525,
+      "grad_norm": 4.139185905456543,
+      "learning_rate": 6.191759999999999e-07,
+      "loss": 0.5245,
+      "step": 95250
+    },
+    {
+      "epoch": 0.953,
+      "grad_norm": 131.5601806640625,
+      "learning_rate": 6.189759999999999e-07,
+      "loss": 0.3855,
+      "step": 95300
+    },
+    {
+      "epoch": 0.9535,
+      "grad_norm": 27.90871238708496,
+      "learning_rate": 6.18776e-07,
+      "loss": 0.4754,
+      "step": 95350
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 82.64385986328125,
+      "learning_rate": 6.18576e-07,
+      "loss": 0.4861,
+      "step": 95400
+    },
+    {
+      "epoch": 0.9545,
+      "grad_norm": 51.01675033569336,
+      "learning_rate": 6.183760000000001e-07,
+      "loss": 0.4428,
+      "step": 95450
+    },
+    {
+      "epoch": 0.955,
+      "grad_norm": 56.236717224121094,
+      "learning_rate": 6.181759999999999e-07,
+      "loss": 0.5291,
+      "step": 95500
+    },
+    {
+      "epoch": 0.9555,
+      "grad_norm": 62.634403228759766,
+      "learning_rate": 6.179759999999999e-07,
+      "loss": 0.4693,
+      "step": 95550
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 29.53840446472168,
+      "learning_rate": 6.17776e-07,
+      "loss": 0.5259,
+      "step": 95600
+    },
+    {
+      "epoch": 0.9565,
+      "grad_norm": 41.582767486572266,
+      "learning_rate": 6.17576e-07,
+      "loss": 0.4315,
+      "step": 95650
+    },
+    {
+      "epoch": 0.957,
+      "grad_norm": 20.49850845336914,
+      "learning_rate": 6.17376e-07,
+      "loss": 0.4245,
+      "step": 95700
+    },
+    {
+      "epoch": 0.9575,
+      "grad_norm": 17.59610939025879,
+      "learning_rate": 6.17176e-07,
+      "loss": 0.4987,
+      "step": 95750
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 80.921630859375,
+      "learning_rate": 6.16976e-07,
+      "loss": 0.3194,
+      "step": 95800
+    },
+    {
+      "epoch": 0.9585,
+      "grad_norm": 84.2488784790039,
+      "learning_rate": 6.167759999999999e-07,
+      "loss": 0.5606,
+      "step": 95850
+    },
+    {
+      "epoch": 0.959,
+      "grad_norm": 0.6134811639785767,
+      "learning_rate": 6.16576e-07,
+      "loss": 0.4759,
+      "step": 95900
+    },
+    {
+      "epoch": 0.9595,
+      "grad_norm": 15.289545059204102,
+      "learning_rate": 6.16376e-07,
+      "loss": 0.5179,
+      "step": 95950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 41.45950698852539,
+      "learning_rate": 6.161759999999999e-07,
+      "loss": 0.6042,
+      "step": 96000
+    },
+    {
+      "epoch": 0.9605,
+      "grad_norm": 2.4903767108917236,
+      "learning_rate": 6.15976e-07,
+      "loss": 0.6048,
+      "step": 96050
+    },
+    {
+      "epoch": 0.961,
+      "grad_norm": 82.29048919677734,
+      "learning_rate": 6.15776e-07,
+      "loss": 0.4689,
+      "step": 96100
+    },
+    {
+      "epoch": 0.9615,
+      "grad_norm": 10.290043830871582,
+      "learning_rate": 6.15576e-07,
+      "loss": 0.4352,
+      "step": 96150
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 56.273529052734375,
+      "learning_rate": 6.15376e-07,
+      "loss": 0.5958,
+      "step": 96200
+    },
+    {
+      "epoch": 0.9625,
+      "grad_norm": 58.597835540771484,
+      "learning_rate": 6.151759999999999e-07,
+      "loss": 0.4868,
+      "step": 96250
+    },
+    {
+      "epoch": 0.963,
+      "grad_norm": 15.38319206237793,
+      "learning_rate": 6.149759999999999e-07,
+      "loss": 0.4837,
+      "step": 96300
+    },
+    {
+      "epoch": 0.9635,
+      "grad_norm": 40.752864837646484,
+      "learning_rate": 6.14776e-07,
+      "loss": 0.4492,
+      "step": 96350
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 7.580194473266602,
+      "learning_rate": 6.14576e-07,
+      "loss": 0.5223,
+      "step": 96400
+    },
+    {
+      "epoch": 0.9645,
+      "grad_norm": 33.51999282836914,
+      "learning_rate": 6.14376e-07,
+      "loss": 0.5388,
+      "step": 96450
+    },
+    {
+      "epoch": 0.965,
+      "grad_norm": 60.9385986328125,
+      "learning_rate": 6.141760000000001e-07,
+      "loss": 0.5853,
+      "step": 96500
+    },
+    {
+      "epoch": 0.9655,
+      "grad_norm": 72.08675384521484,
+      "learning_rate": 6.139759999999999e-07,
+      "loss": 0.567,
+      "step": 96550
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 27.145662307739258,
+      "learning_rate": 6.137759999999999e-07,
+      "loss": 0.5532,
+      "step": 96600
+    },
+    {
+      "epoch": 0.9665,
+      "grad_norm": 91.9208984375,
+      "learning_rate": 6.1358e-07,
+      "loss": 0.4526,
+      "step": 96650
+    },
+    {
+      "epoch": 0.967,
+      "grad_norm": 6.773651123046875,
+      "learning_rate": 6.1338e-07,
+      "loss": 0.467,
+      "step": 96700
+    },
+    {
+      "epoch": 0.9675,
+      "grad_norm": 54.172584533691406,
+      "learning_rate": 6.131799999999999e-07,
+      "loss": 0.5223,
+      "step": 96750
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 23.472658157348633,
+      "learning_rate": 6.1298e-07,
+      "loss": 0.4758,
+      "step": 96800
+    },
+    {
+      "epoch": 0.9685,
+      "grad_norm": 104.0491714477539,
+      "learning_rate": 6.1278e-07,
+      "loss": 0.5201,
+      "step": 96850
+    },
+    {
+      "epoch": 0.969,
+      "grad_norm": 11.634442329406738,
+      "learning_rate": 6.125799999999999e-07,
+      "loss": 0.4486,
+      "step": 96900
+    },
+    {
+      "epoch": 0.9695,
+      "grad_norm": 24.342527389526367,
+      "learning_rate": 6.1238e-07,
+      "loss": 0.5842,
+      "step": 96950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.33423900604248,
+      "learning_rate": 6.121799999999999e-07,
+      "loss": 0.4822,
+      "step": 97000
+    },
+    {
+      "epoch": 0.9705,
+      "grad_norm": 73.634033203125,
+      "learning_rate": 6.119799999999999e-07,
+      "loss": 0.4194,
+      "step": 97050
+    },
+    {
+      "epoch": 0.971,
+      "grad_norm": 90.969970703125,
+      "learning_rate": 6.1178e-07,
+      "loss": 0.6029,
+      "step": 97100
+    },
+    {
+      "epoch": 0.9715,
+      "grad_norm": 17.80976104736328,
+      "learning_rate": 6.1158e-07,
+      "loss": 0.4477,
+      "step": 97150
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 81.571533203125,
+      "learning_rate": 6.1138e-07,
+      "loss": 0.7005,
+      "step": 97200
+    },
+    {
+      "epoch": 0.9725,
+      "grad_norm": 41.37546157836914,
+      "learning_rate": 6.111799999999999e-07,
+      "loss": 0.5194,
+      "step": 97250
+    },
+    {
+      "epoch": 0.973,
+      "grad_norm": 100.39408874511719,
+      "learning_rate": 6.109799999999999e-07,
+      "loss": 0.3624,
+      "step": 97300
+    },
+    {
+      "epoch": 0.9735,
+      "grad_norm": 65.12748718261719,
+      "learning_rate": 6.107799999999999e-07,
+      "loss": 0.5751,
+      "step": 97350
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 33.728546142578125,
+      "learning_rate": 6.1058e-07,
+      "loss": 0.521,
+      "step": 97400
+    },
+    {
+      "epoch": 0.9745,
+      "grad_norm": 69.31006622314453,
+      "learning_rate": 6.10384e-07,
+      "loss": 0.4695,
+      "step": 97450
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 38.2792854309082,
+      "learning_rate": 6.101839999999999e-07,
+      "loss": 0.3575,
+      "step": 97500
+    },
+    {
+      "epoch": 0.9755,
+      "grad_norm": 69.43871307373047,
+      "learning_rate": 6.09984e-07,
+      "loss": 0.3759,
+      "step": 97550
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 68.89128112792969,
+      "learning_rate": 6.09784e-07,
+      "loss": 0.4355,
+      "step": 97600
+    },
+    {
+      "epoch": 0.9765,
+      "grad_norm": 41.74790573120117,
+      "learning_rate": 6.095839999999999e-07,
+      "loss": 0.6918,
+      "step": 97650
+    },
+    {
+      "epoch": 0.977,
+      "grad_norm": 12.850433349609375,
+      "learning_rate": 6.09384e-07,
+      "loss": 0.556,
+      "step": 97700
+    },
+    {
+      "epoch": 0.9775,
+      "grad_norm": 43.75520324707031,
+      "learning_rate": 6.091839999999999e-07,
+      "loss": 0.4272,
+      "step": 97750
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 9.665956497192383,
+      "learning_rate": 6.089839999999999e-07,
+      "loss": 0.4407,
+      "step": 97800
+    },
+    {
+      "epoch": 0.9785,
+      "grad_norm": 22.188209533691406,
+      "learning_rate": 6.08784e-07,
+      "loss": 0.4733,
+      "step": 97850
+    },
+    {
+      "epoch": 0.979,
+      "grad_norm": 108.53839111328125,
+      "learning_rate": 6.08584e-07,
+      "loss": 0.5335,
+      "step": 97900
+    },
+    {
+      "epoch": 0.9795,
+      "grad_norm": 58.352508544921875,
+      "learning_rate": 6.083840000000001e-07,
+      "loss": 0.478,
+      "step": 97950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 57.22300338745117,
+      "learning_rate": 6.081839999999999e-07,
+      "loss": 0.543,
+      "step": 98000
+    },
+    {
+      "epoch": 0.9805,
+      "grad_norm": 42.58542251586914,
+      "learning_rate": 6.079839999999999e-07,
+      "loss": 0.4557,
+      "step": 98050
+    },
+    {
+      "epoch": 0.981,
+      "grad_norm": 73.4317626953125,
+      "learning_rate": 6.07784e-07,
+      "loss": 0.4937,
+      "step": 98100
+    },
+    {
+      "epoch": 0.9815,
+      "grad_norm": 24.243736267089844,
+      "learning_rate": 6.07584e-07,
+      "loss": 0.4762,
+      "step": 98150
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 0.41077935695648193,
+      "learning_rate": 6.07384e-07,
+      "loss": 0.5536,
+      "step": 98200
+    },
+    {
+      "epoch": 0.9825,
+      "grad_norm": 3.0135037899017334,
+      "learning_rate": 6.07184e-07,
+      "loss": 0.3889,
+      "step": 98250
+    },
+    {
+      "epoch": 0.983,
+      "grad_norm": 27.06598472595215,
+      "learning_rate": 6.06984e-07,
+      "loss": 0.5288,
+      "step": 98300
+    },
+    {
+      "epoch": 0.9835,
+      "grad_norm": 79.37501525878906,
+      "learning_rate": 6.067839999999999e-07,
+      "loss": 0.5145,
+      "step": 98350
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 10.967093467712402,
+      "learning_rate": 6.06584e-07,
+      "loss": 0.5057,
+      "step": 98400
+    },
+    {
+      "epoch": 0.9845,
+      "grad_norm": 86.88298797607422,
+      "learning_rate": 6.06384e-07,
+      "loss": 0.4921,
+      "step": 98450
+    },
+    {
+      "epoch": 0.985,
+      "grad_norm": 83.58354187011719,
+      "learning_rate": 6.061839999999999e-07,
+      "loss": 0.6873,
+      "step": 98500
+    },
+    {
+      "epoch": 0.9855,
+      "grad_norm": 54.29396057128906,
+      "learning_rate": 6.05984e-07,
+      "loss": 0.5628,
+      "step": 98550
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 93.75994873046875,
+      "learning_rate": 6.05784e-07,
+      "loss": 0.4604,
+      "step": 98600
+    },
+    {
+      "epoch": 0.9865,
+      "grad_norm": 26.21271324157715,
+      "learning_rate": 6.05584e-07,
+      "loss": 0.47,
+      "step": 98650
+    },
+    {
+      "epoch": 0.987,
+      "grad_norm": 75.35508728027344,
+      "learning_rate": 6.05384e-07,
+      "loss": 0.4449,
+      "step": 98700
+    },
+    {
+      "epoch": 0.9875,
+      "grad_norm": 34.0965690612793,
+      "learning_rate": 6.05184e-07,
+      "loss": 0.4985,
+      "step": 98750
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 15.451560974121094,
+      "learning_rate": 6.049839999999999e-07,
+      "loss": 0.4956,
+      "step": 98800
+    },
+    {
+      "epoch": 0.9885,
+      "grad_norm": 6.660648822784424,
+      "learning_rate": 6.04784e-07,
+      "loss": 0.3919,
+      "step": 98850
+    },
+    {
+      "epoch": 0.989,
+      "grad_norm": 39.9170036315918,
+      "learning_rate": 6.04584e-07,
+      "loss": 0.497,
+      "step": 98900
+    },
+    {
+      "epoch": 0.9895,
+      "grad_norm": 33.44998550415039,
+      "learning_rate": 6.04384e-07,
+      "loss": 0.6224,
+      "step": 98950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 28.230859756469727,
+      "learning_rate": 6.041840000000001e-07,
+      "loss": 0.4241,
+      "step": 99000
+    },
+    {
+      "epoch": 0.9905,
+      "grad_norm": 104.45681762695312,
+      "learning_rate": 6.039839999999999e-07,
+      "loss": 0.4589,
+      "step": 99050
+    },
+    {
+      "epoch": 0.991,
+      "grad_norm": 8.24346923828125,
+      "learning_rate": 6.037839999999999e-07,
+      "loss": 0.4088,
+      "step": 99100
+    },
+    {
+      "epoch": 0.9915,
+      "grad_norm": 73.47013092041016,
+      "learning_rate": 6.03584e-07,
+      "loss": 0.4545,
+      "step": 99150
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 88.34424591064453,
+      "learning_rate": 6.03384e-07,
+      "loss": 0.4967,
+      "step": 99200
+    },
+    {
+      "epoch": 0.9925,
+      "grad_norm": 19.245553970336914,
+      "learning_rate": 6.03184e-07,
+      "loss": 0.3781,
+      "step": 99250
+    },
+    {
+      "epoch": 0.993,
+      "grad_norm": 24.192604064941406,
+      "learning_rate": 6.02984e-07,
+      "loss": 0.5745,
+      "step": 99300
+    },
+    {
+      "epoch": 0.9935,
+      "grad_norm": 30.267501831054688,
+      "learning_rate": 6.02784e-07,
+      "loss": 0.4486,
+      "step": 99350
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 102.20484161376953,
+      "learning_rate": 6.025839999999999e-07,
+      "loss": 0.5407,
+      "step": 99400
+    },
+    {
+      "epoch": 0.9945,
+      "grad_norm": 65.89080810546875,
+      "learning_rate": 6.02384e-07,
+      "loss": 0.5351,
+      "step": 99450
+    },
+    {
+      "epoch": 0.995,
+      "grad_norm": 84.1611557006836,
+      "learning_rate": 6.02184e-07,
+      "loss": 0.4555,
+      "step": 99500
+    },
+    {
+      "epoch": 0.9955,
+      "grad_norm": 4.047856330871582,
+      "learning_rate": 6.019839999999999e-07,
+      "loss": 0.4759,
+      "step": 99550
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 8.221104621887207,
+      "learning_rate": 6.01784e-07,
+      "loss": 0.3095,
+      "step": 99600
+    },
+    {
+      "epoch": 0.9965,
+      "grad_norm": 56.98789978027344,
+      "learning_rate": 6.01584e-07,
+      "loss": 0.4643,
+      "step": 99650
+    },
+    {
+      "epoch": 0.997,
+      "grad_norm": 12.39257526397705,
+      "learning_rate": 6.01384e-07,
+      "loss": 0.3251,
+      "step": 99700
+    },
+    {
+      "epoch": 0.9975,
+      "grad_norm": 20.75105094909668,
+      "learning_rate": 6.01184e-07,
+      "loss": 0.5136,
+      "step": 99750
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 47.02289962768555,
+      "learning_rate": 6.009839999999999e-07,
+      "loss": 0.4591,
+      "step": 99800
+    },
+    {
+      "epoch": 0.9985,
+      "grad_norm": 97.42501831054688,
+      "learning_rate": 6.007839999999999e-07,
+      "loss": 0.5596,
+      "step": 99850
+    },
+    {
+      "epoch": 0.999,
+      "grad_norm": 66.5621566772461,
+      "learning_rate": 6.00584e-07,
+      "loss": 0.4794,
+      "step": 99900
+    },
+    {
+      "epoch": 0.9995,
+      "grad_norm": 9.777603149414062,
+      "learning_rate": 6.00384e-07,
+      "loss": 0.4818,
+      "step": 99950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 66.84056854248047,
+      "learning_rate": 6.001840000000001e-07,
+      "loss": 0.4902,
+      "step": 100000
+    },
+    {
+      "epoch": 1.0005,
+      "grad_norm": 0.26840323209762573,
+      "learning_rate": 5.99984e-07,
+      "loss": 0.4989,
+      "step": 100050
+    },
+    {
+      "epoch": 1.001,
+      "grad_norm": 1.4899623394012451,
+      "learning_rate": 5.997839999999999e-07,
+      "loss": 0.466,
+      "step": 100100
+    },
+    {
+      "epoch": 1.0015,
+      "grad_norm": 56.374874114990234,
+      "learning_rate": 5.99584e-07,
+      "loss": 0.5739,
+      "step": 100150
+    },
+    {
+      "epoch": 1.002,
+      "grad_norm": 28.144792556762695,
+      "learning_rate": 5.99384e-07,
+      "loss": 0.5244,
+      "step": 100200
+    },
+    {
+      "epoch": 1.0025,
+      "grad_norm": 41.527069091796875,
+      "learning_rate": 5.99184e-07,
+      "loss": 0.4076,
+      "step": 100250
+    },
+    {
+      "epoch": 1.003,
+      "grad_norm": 24.504615783691406,
+      "learning_rate": 5.98984e-07,
+      "loss": 0.4945,
+      "step": 100300
+    },
+    {
+      "epoch": 1.0035,
+      "grad_norm": 59.61265563964844,
+      "learning_rate": 5.98784e-07,
+      "loss": 0.5471,
+      "step": 100350
+    },
+    {
+      "epoch": 1.004,
+      "grad_norm": 4.118966102600098,
+      "learning_rate": 5.98584e-07,
+      "loss": 0.5698,
+      "step": 100400
+    },
+    {
+      "epoch": 1.0045,
+      "grad_norm": 83.80474853515625,
+      "learning_rate": 5.98384e-07,
+      "loss": 0.4028,
+      "step": 100450
+    },
+    {
+      "epoch": 1.005,
+      "grad_norm": 82.36299133300781,
+      "learning_rate": 5.98184e-07,
+      "loss": 0.3842,
+      "step": 100500
+    },
+    {
+      "epoch": 1.0055,
+      "grad_norm": 26.00364112854004,
+      "learning_rate": 5.979839999999999e-07,
+      "loss": 0.4643,
+      "step": 100550
+    },
+    {
+      "epoch": 1.006,
+      "grad_norm": 3.6861441135406494,
+      "learning_rate": 5.97784e-07,
+      "loss": 0.5716,
+      "step": 100600
+    },
+    {
+      "epoch": 1.0065,
+      "grad_norm": 9.791892051696777,
+      "learning_rate": 5.97584e-07,
+      "loss": 0.4626,
+      "step": 100650
+    },
+    {
+      "epoch": 1.007,
+      "grad_norm": 32.55017852783203,
+      "learning_rate": 5.97384e-07,
+      "loss": 0.3589,
+      "step": 100700
+    },
+    {
+      "epoch": 1.0075,
+      "grad_norm": 102.46571350097656,
+      "learning_rate": 5.971840000000001e-07,
+      "loss": 0.5368,
+      "step": 100750
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 3.5553598403930664,
+      "learning_rate": 5.969839999999999e-07,
+      "loss": 0.4756,
+      "step": 100800
+    },
+    {
+      "epoch": 1.0085,
+      "grad_norm": 84.53421020507812,
+      "learning_rate": 5.967839999999999e-07,
+      "loss": 0.4855,
+      "step": 100850
+    },
+    {
+      "epoch": 1.009,
+      "grad_norm": 50.297672271728516,
+      "learning_rate": 5.96584e-07,
+      "loss": 0.4461,
+      "step": 100900
+    },
+    {
+      "epoch": 1.0095,
+      "grad_norm": 46.95347213745117,
+      "learning_rate": 5.96384e-07,
+      "loss": 0.2661,
+      "step": 100950
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 46.71750259399414,
+      "learning_rate": 5.96184e-07,
+      "loss": 0.5056,
+      "step": 101000
+    },
+    {
+      "epoch": 1.0105,
+      "grad_norm": 65.1976547241211,
+      "learning_rate": 5.95984e-07,
+      "loss": 0.4613,
+      "step": 101050
+    },
+    {
+      "epoch": 1.011,
+      "grad_norm": 53.80318069458008,
+      "learning_rate": 5.95784e-07,
+      "loss": 0.4472,
+      "step": 101100
+    },
+    {
+      "epoch": 1.0115,
+      "grad_norm": 82.39476776123047,
+      "learning_rate": 5.955839999999999e-07,
+      "loss": 0.5717,
+      "step": 101150
+    },
+    {
+      "epoch": 1.012,
+      "grad_norm": 39.730159759521484,
+      "learning_rate": 5.95384e-07,
+      "loss": 0.4549,
+      "step": 101200
+    },
+    {
+      "epoch": 1.0125,
+      "grad_norm": 106.95108795166016,
+      "learning_rate": 5.95184e-07,
+      "loss": 0.489,
+      "step": 101250
+    },
+    {
+      "epoch": 1.013,
+      "grad_norm": 48.25342559814453,
+      "learning_rate": 5.949839999999999e-07,
+      "loss": 0.4537,
+      "step": 101300
+    },
+    {
+      "epoch": 1.0135,
+      "grad_norm": 2.1384692192077637,
+      "learning_rate": 5.94784e-07,
+      "loss": 0.6137,
+      "step": 101350
+    },
+    {
+      "epoch": 1.014,
+      "grad_norm": 59.659515380859375,
+      "learning_rate": 5.94584e-07,
+      "loss": 0.4321,
+      "step": 101400
+    },
+    {
+      "epoch": 1.0145,
+      "grad_norm": 60.75330352783203,
+      "learning_rate": 5.94384e-07,
+      "loss": 0.4137,
+      "step": 101450
+    },
+    {
+      "epoch": 1.015,
+      "grad_norm": 84.83941650390625,
+      "learning_rate": 5.94184e-07,
+      "loss": 0.7533,
+      "step": 101500
+    },
+    {
+      "epoch": 1.0155,
+      "grad_norm": 13.40371036529541,
+      "learning_rate": 5.939839999999999e-07,
+      "loss": 0.4591,
+      "step": 101550
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 38.117313385009766,
+      "learning_rate": 5.937839999999999e-07,
+      "loss": 0.5045,
+      "step": 101600
+    },
+    {
+      "epoch": 1.0165,
+      "grad_norm": 1.0856471061706543,
+      "learning_rate": 5.93584e-07,
+      "loss": 0.3855,
+      "step": 101650
+    },
+    {
+      "epoch": 1.017,
+      "grad_norm": 1.0429091453552246,
+      "learning_rate": 5.93384e-07,
+      "loss": 0.4512,
+      "step": 101700
+    },
+    {
+      "epoch": 1.0175,
+      "grad_norm": 19.322357177734375,
+      "learning_rate": 5.93184e-07,
+      "loss": 0.3774,
+      "step": 101750
+    },
+    {
+      "epoch": 1.018,
+      "grad_norm": 104.85911560058594,
+      "learning_rate": 5.92984e-07,
+      "loss": 0.4906,
+      "step": 101800
+    },
+    {
+      "epoch": 1.0185,
+      "grad_norm": 48.950130462646484,
+      "learning_rate": 5.927839999999999e-07,
+      "loss": 0.4659,
+      "step": 101850
+    },
+    {
+      "epoch": 1.019,
+      "grad_norm": 40.40967559814453,
+      "learning_rate": 5.925839999999999e-07,
+      "loss": 0.4826,
+      "step": 101900
+    },
+    {
+      "epoch": 1.0195,
+      "grad_norm": 32.901214599609375,
+      "learning_rate": 5.92384e-07,
+      "loss": 0.3543,
+      "step": 101950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 22.917282104492188,
+      "learning_rate": 5.92184e-07,
+      "loss": 0.3971,
+      "step": 102000
+    },
+    {
+      "epoch": 1.0205,
+      "grad_norm": 70.27293395996094,
+      "learning_rate": 5.919839999999999e-07,
+      "loss": 0.4203,
+      "step": 102050
+    },
+    {
+      "epoch": 1.021,
+      "grad_norm": 58.08955001831055,
+      "learning_rate": 5.91784e-07,
+      "loss": 0.4806,
+      "step": 102100
+    },
+    {
+      "epoch": 1.0215,
+      "grad_norm": 0.6320064663887024,
+      "learning_rate": 5.91584e-07,
+      "loss": 0.4064,
+      "step": 102150
+    },
+    {
+      "epoch": 1.022,
+      "grad_norm": 13.316550254821777,
+      "learning_rate": 5.913839999999999e-07,
+      "loss": 0.4039,
+      "step": 102200
+    },
+    {
+      "epoch": 1.0225,
+      "grad_norm": 86.25184631347656,
+      "learning_rate": 5.91184e-07,
+      "loss": 0.389,
+      "step": 102250
+    },
+    {
+      "epoch": 1.023,
+      "grad_norm": 11.370412826538086,
+      "learning_rate": 5.909839999999999e-07,
+      "loss": 0.4349,
+      "step": 102300
+    },
+    {
+      "epoch": 1.0235,
+      "grad_norm": 7.496046543121338,
+      "learning_rate": 5.907839999999999e-07,
+      "loss": 0.4676,
+      "step": 102350
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 15.750993728637695,
+      "learning_rate": 5.90584e-07,
+      "loss": 0.5462,
+      "step": 102400
+    },
+    {
+      "epoch": 1.0245,
+      "grad_norm": 108.76589965820312,
+      "learning_rate": 5.90384e-07,
+      "loss": 0.4425,
+      "step": 102450
+    },
+    {
+      "epoch": 1.025,
+      "grad_norm": 85.9919204711914,
+      "learning_rate": 5.901840000000001e-07,
+      "loss": 0.4755,
+      "step": 102500
+    },
+    {
+      "epoch": 1.0255,
+      "grad_norm": 84.9974594116211,
+      "learning_rate": 5.899839999999999e-07,
+      "loss": 0.3698,
+      "step": 102550
+    },
+    {
+      "epoch": 1.026,
+      "grad_norm": 53.898643493652344,
+      "learning_rate": 5.897839999999999e-07,
+      "loss": 0.5112,
+      "step": 102600
+    },
+    {
+      "epoch": 1.0265,
+      "grad_norm": 50.896507263183594,
+      "learning_rate": 5.89584e-07,
+      "loss": 0.3905,
+      "step": 102650
+    },
+    {
+      "epoch": 1.027,
+      "grad_norm": 78.91048431396484,
+      "learning_rate": 5.89384e-07,
+      "loss": 0.4481,
+      "step": 102700
+    },
+    {
+      "epoch": 1.0275,
+      "grad_norm": 82.20804595947266,
+      "learning_rate": 5.89184e-07,
+      "loss": 0.4744,
+      "step": 102750
+    },
+    {
+      "epoch": 1.028,
+      "grad_norm": 72.1124038696289,
+      "learning_rate": 5.88984e-07,
+      "loss": 0.5971,
+      "step": 102800
+    },
+    {
+      "epoch": 1.0285,
+      "grad_norm": 6.744050979614258,
+      "learning_rate": 5.88784e-07,
+      "loss": 0.4036,
+      "step": 102850
+    },
+    {
+      "epoch": 1.029,
+      "grad_norm": 102.9438247680664,
+      "learning_rate": 5.885839999999999e-07,
+      "loss": 0.3638,
+      "step": 102900
+    },
+    {
+      "epoch": 1.0295,
+      "grad_norm": 91.29195404052734,
+      "learning_rate": 5.88384e-07,
+      "loss": 0.4288,
+      "step": 102950
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.420190811157227,
+      "learning_rate": 5.88184e-07,
+      "loss": 0.5381,
+      "step": 103000
+    },
+    {
+      "epoch": 1.0305,
+      "grad_norm": 3.8758227825164795,
+      "learning_rate": 5.879839999999999e-07,
+      "loss": 0.4904,
+      "step": 103050
+    },
+    {
+      "epoch": 1.031,
+      "grad_norm": 42.36983871459961,
+      "learning_rate": 5.87784e-07,
+      "loss": 0.4687,
+      "step": 103100
+    },
+    {
+      "epoch": 1.0315,
+      "grad_norm": 4.480782508850098,
+      "learning_rate": 5.87584e-07,
+      "loss": 0.3783,
+      "step": 103150
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 16.480192184448242,
+      "learning_rate": 5.87384e-07,
+      "loss": 0.4104,
+      "step": 103200
+    },
+    {
+      "epoch": 1.0325,
+      "grad_norm": 56.70964050292969,
+      "learning_rate": 5.87184e-07,
+      "loss": 0.4744,
+      "step": 103250
+    },
+    {
+      "epoch": 1.033,
+      "grad_norm": 79.56087493896484,
+      "learning_rate": 5.869839999999999e-07,
+      "loss": 0.4447,
+      "step": 103300
+    },
+    {
+      "epoch": 1.0335,
+      "grad_norm": 60.962284088134766,
+      "learning_rate": 5.867839999999999e-07,
+      "loss": 0.5019,
+      "step": 103350
+    },
+    {
+      "epoch": 1.034,
+      "grad_norm": 58.27140808105469,
+      "learning_rate": 5.86584e-07,
+      "loss": 0.5927,
+      "step": 103400
+    },
+    {
+      "epoch": 1.0345,
+      "grad_norm": 39.9395637512207,
+      "learning_rate": 5.86384e-07,
+      "loss": 0.4556,
+      "step": 103450
+    },
+    {
+      "epoch": 1.035,
+      "grad_norm": 10.879745483398438,
+      "learning_rate": 5.86184e-07,
+      "loss": 0.5386,
+      "step": 103500
+    },
+    {
+      "epoch": 1.0355,
+      "grad_norm": 14.985457420349121,
+      "learning_rate": 5.85984e-07,
+      "loss": 0.4561,
+      "step": 103550
+    },
+    {
+      "epoch": 1.036,
+      "grad_norm": 0.901373565196991,
+      "learning_rate": 5.85788e-07,
+      "loss": 0.486,
+      "step": 103600
+    },
+    {
+      "epoch": 1.0365,
+      "grad_norm": 76.16960906982422,
+      "learning_rate": 5.855879999999999e-07,
+      "loss": 0.5131,
+      "step": 103650
+    },
+    {
+      "epoch": 1.037,
+      "grad_norm": 71.22393798828125,
+      "learning_rate": 5.85388e-07,
+      "loss": 0.6253,
+      "step": 103700
+    },
+    {
+      "epoch": 1.0375,
+      "grad_norm": 72.47522735595703,
+      "learning_rate": 5.85188e-07,
+      "loss": 0.4038,
+      "step": 103750
+    },
+    {
+      "epoch": 1.038,
+      "grad_norm": 35.06593322753906,
+      "learning_rate": 5.849879999999999e-07,
+      "loss": 0.4235,
+      "step": 103800
+    },
+    {
+      "epoch": 1.0385,
+      "grad_norm": 58.12506866455078,
+      "learning_rate": 5.84788e-07,
+      "loss": 0.4734,
+      "step": 103850
+    },
+    {
+      "epoch": 1.039,
+      "grad_norm": 70.32412719726562,
+      "learning_rate": 5.84588e-07,
+      "loss": 0.4489,
+      "step": 103900
+    },
+    {
+      "epoch": 1.0395,
+      "grad_norm": 94.34310150146484,
+      "learning_rate": 5.84388e-07,
+      "loss": 0.5206,
+      "step": 103950
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 96.84242248535156,
+      "learning_rate": 5.84188e-07,
+      "loss": 0.4214,
+      "step": 104000
+    },
+    {
+      "epoch": 1.0405,
+      "grad_norm": 50.48087692260742,
+      "learning_rate": 5.839879999999999e-07,
+      "loss": 0.4325,
+      "step": 104050
+    },
+    {
+      "epoch": 1.041,
+      "grad_norm": 15.453256607055664,
+      "learning_rate": 5.837879999999999e-07,
+      "loss": 0.4099,
+      "step": 104100
+    },
+    {
+      "epoch": 1.0415,
+      "grad_norm": 7.770763397216797,
+      "learning_rate": 5.83588e-07,
+      "loss": 0.5572,
+      "step": 104150
+    },
+    {
+      "epoch": 1.042,
+      "grad_norm": 75.37487030029297,
+      "learning_rate": 5.83388e-07,
+      "loss": 0.6112,
+      "step": 104200
+    },
+    {
+      "epoch": 1.0425,
+      "grad_norm": 3.0867295265197754,
+      "learning_rate": 5.83188e-07,
+      "loss": 0.4438,
+      "step": 104250
+    },
+    {
+      "epoch": 1.043,
+      "grad_norm": 82.33965301513672,
+      "learning_rate": 5.82988e-07,
+      "loss": 0.5488,
+      "step": 104300
+    },
+    {
+      "epoch": 1.0435,
+      "grad_norm": 108.06182861328125,
+      "learning_rate": 5.827879999999999e-07,
+      "loss": 0.4588,
+      "step": 104350
+    },
+    {
+      "epoch": 1.044,
+      "grad_norm": 7.8545660972595215,
+      "learning_rate": 5.825879999999999e-07,
+      "loss": 0.4324,
+      "step": 104400
+    },
+    {
+      "epoch": 1.0445,
+      "grad_norm": 105.07747650146484,
+      "learning_rate": 5.82388e-07,
+      "loss": 0.4492,
+      "step": 104450
+    },
+    {
+      "epoch": 1.045,
+      "grad_norm": 50.81571578979492,
+      "learning_rate": 5.82188e-07,
+      "loss": 0.6157,
+      "step": 104500
+    },
+    {
+      "epoch": 1.0455,
+      "grad_norm": 102.53087615966797,
+      "learning_rate": 5.819879999999999e-07,
+      "loss": 0.3951,
+      "step": 104550
+    },
+    {
+      "epoch": 1.046,
+      "grad_norm": 40.119380950927734,
+      "learning_rate": 5.81788e-07,
+      "loss": 0.4466,
+      "step": 104600
+    },
+    {
+      "epoch": 1.0465,
+      "grad_norm": 98.8756103515625,
+      "learning_rate": 5.81588e-07,
+      "loss": 0.4351,
+      "step": 104650
+    },
+    {
+      "epoch": 1.047,
+      "grad_norm": 7.848785400390625,
+      "learning_rate": 5.813879999999999e-07,
+      "loss": 0.4286,
+      "step": 104700
+    },
+    {
+      "epoch": 1.0475,
+      "grad_norm": 57.9787712097168,
+      "learning_rate": 5.81188e-07,
+      "loss": 0.4914,
+      "step": 104750
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 34.086063385009766,
+      "learning_rate": 5.809879999999999e-07,
+      "loss": 0.5624,
+      "step": 104800
+    },
+    {
+      "epoch": 1.0485,
+      "grad_norm": 38.03075408935547,
+      "learning_rate": 5.807879999999999e-07,
+      "loss": 0.4121,
+      "step": 104850
+    },
+    {
+      "epoch": 1.049,
+      "grad_norm": 113.10247039794922,
+      "learning_rate": 5.80588e-07,
+      "loss": 0.5234,
+      "step": 104900
+    },
+    {
+      "epoch": 1.0495,
+      "grad_norm": 67.31755828857422,
+      "learning_rate": 5.80388e-07,
+      "loss": 0.4457,
+      "step": 104950
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.373214244842529,
+      "learning_rate": 5.801880000000001e-07,
+      "loss": 0.4674,
+      "step": 105000
+    },
+    {
+      "epoch": 1.0505,
+      "grad_norm": 24.029144287109375,
+      "learning_rate": 5.799879999999999e-07,
+      "loss": 0.4393,
+      "step": 105050
+    },
+    {
+      "epoch": 1.051,
+      "grad_norm": 76.45175170898438,
+      "learning_rate": 5.797879999999999e-07,
+      "loss": 0.6139,
+      "step": 105100
+    },
+    {
+      "epoch": 1.0515,
+      "grad_norm": 90.8389663696289,
+      "learning_rate": 5.79588e-07,
+      "loss": 0.4096,
+      "step": 105150
+    },
+    {
+      "epoch": 1.052,
+      "grad_norm": 52.22407531738281,
+      "learning_rate": 5.79388e-07,
+      "loss": 0.4983,
+      "step": 105200
+    },
+    {
+      "epoch": 1.0525,
+      "grad_norm": 28.466354370117188,
+      "learning_rate": 5.79188e-07,
+      "loss": 0.4473,
+      "step": 105250
+    },
+    {
+      "epoch": 1.053,
+      "grad_norm": 74.26171112060547,
+      "learning_rate": 5.78988e-07,
+      "loss": 0.5071,
+      "step": 105300
+    },
+    {
+      "epoch": 1.0535,
+      "grad_norm": 45.36643981933594,
+      "learning_rate": 5.78788e-07,
+      "loss": 0.4621,
+      "step": 105350
+    },
+    {
+      "epoch": 1.054,
+      "grad_norm": 22.525741577148438,
+      "learning_rate": 5.785879999999999e-07,
+      "loss": 0.5238,
+      "step": 105400
+    },
+    {
+      "epoch": 1.0545,
+      "grad_norm": 25.078907012939453,
+      "learning_rate": 5.78388e-07,
+      "loss": 0.472,
+      "step": 105450
+    },
+    {
+      "epoch": 1.055,
+      "grad_norm": 29.455020904541016,
+      "learning_rate": 5.78188e-07,
+      "loss": 0.3387,
+      "step": 105500
+    },
+    {
+      "epoch": 1.0555,
+      "grad_norm": 4.957467555999756,
+      "learning_rate": 5.779879999999999e-07,
+      "loss": 0.4357,
+      "step": 105550
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 56.372745513916016,
+      "learning_rate": 5.77788e-07,
+      "loss": 0.5183,
+      "step": 105600
+    },
+    {
+      "epoch": 1.0565,
+      "grad_norm": 1.047702670097351,
+      "learning_rate": 5.77588e-07,
+      "loss": 0.3755,
+      "step": 105650
+    },
+    {
+      "epoch": 1.057,
+      "grad_norm": 26.11679458618164,
+      "learning_rate": 5.77388e-07,
+      "loss": 0.431,
+      "step": 105700
+    },
+    {
+      "epoch": 1.0575,
+      "grad_norm": 81.7731704711914,
+      "learning_rate": 5.77192e-07,
+      "loss": 0.642,
+      "step": 105750
+    },
+    {
+      "epoch": 1.058,
+      "grad_norm": 2.761612892150879,
+      "learning_rate": 5.769919999999999e-07,
+      "loss": 0.4243,
+      "step": 105800
+    },
+    {
+      "epoch": 1.0585,
+      "grad_norm": 16.649229049682617,
+      "learning_rate": 5.767919999999999e-07,
+      "loss": 0.4652,
+      "step": 105850
+    },
+    {
+      "epoch": 1.059,
+      "grad_norm": 74.75067138671875,
+      "learning_rate": 5.76592e-07,
+      "loss": 0.4756,
+      "step": 105900
+    },
+    {
+      "epoch": 1.0594999999999999,
+      "grad_norm": 6.984798431396484,
+      "learning_rate": 5.76392e-07,
+      "loss": 0.4214,
+      "step": 105950
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 40.673606872558594,
+      "learning_rate": 5.76192e-07,
+      "loss": 0.5516,
+      "step": 106000
+    },
+    {
+      "epoch": 1.0605,
+      "grad_norm": 111.4358901977539,
+      "learning_rate": 5.75992e-07,
+      "loss": 0.4571,
+      "step": 106050
+    },
+    {
+      "epoch": 1.061,
+      "grad_norm": 22.08611297607422,
+      "learning_rate": 5.757919999999999e-07,
+      "loss": 0.4519,
+      "step": 106100
+    },
+    {
+      "epoch": 1.0615,
+      "grad_norm": 6.3416666984558105,
+      "learning_rate": 5.755919999999999e-07,
+      "loss": 0.3882,
+      "step": 106150
+    },
+    {
+      "epoch": 1.062,
+      "grad_norm": 18.259502410888672,
+      "learning_rate": 5.75392e-07,
+      "loss": 0.3961,
+      "step": 106200
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 76.18441772460938,
+      "learning_rate": 5.75192e-07,
+      "loss": 0.4029,
+      "step": 106250
+    },
+    {
+      "epoch": 1.063,
+      "grad_norm": 47.05423355102539,
+      "learning_rate": 5.749919999999999e-07,
+      "loss": 0.3685,
+      "step": 106300
+    },
+    {
+      "epoch": 1.0635,
+      "grad_norm": 69.71207427978516,
+      "learning_rate": 5.74792e-07,
+      "loss": 0.4922,
+      "step": 106350
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 32.429664611816406,
+      "learning_rate": 5.74592e-07,
+      "loss": 0.6055,
+      "step": 106400
+    },
+    {
+      "epoch": 1.0645,
+      "grad_norm": 105.841552734375,
+      "learning_rate": 5.743919999999999e-07,
+      "loss": 0.4959,
+      "step": 106450
+    },
+    {
+      "epoch": 1.065,
+      "grad_norm": 65.44140625,
+      "learning_rate": 5.74192e-07,
+      "loss": 0.4779,
+      "step": 106500
+    },
+    {
+      "epoch": 1.0655000000000001,
+      "grad_norm": 31.591449737548828,
+      "learning_rate": 5.739919999999999e-07,
+      "loss": 0.5328,
+      "step": 106550
+    },
+    {
+      "epoch": 1.066,
+      "grad_norm": 39.16038513183594,
+      "learning_rate": 5.737919999999999e-07,
+      "loss": 0.4452,
+      "step": 106600
+    },
+    {
+      "epoch": 1.0665,
+      "grad_norm": 0.7136451601982117,
+      "learning_rate": 5.73592e-07,
+      "loss": 0.4545,
+      "step": 106650
+    },
+    {
+      "epoch": 1.067,
+      "grad_norm": 78.04670715332031,
+      "learning_rate": 5.73392e-07,
+      "loss": 0.5004,
+      "step": 106700
+    },
+    {
+      "epoch": 1.0675,
+      "grad_norm": 77.70978546142578,
+      "learning_rate": 5.73192e-07,
+      "loss": 0.4688,
+      "step": 106750
+    },
+    {
+      "epoch": 1.068,
+      "grad_norm": 2.930210590362549,
+      "learning_rate": 5.729919999999999e-07,
+      "loss": 0.504,
+      "step": 106800
+    },
+    {
+      "epoch": 1.0685,
+      "grad_norm": 2.333472967147827,
+      "learning_rate": 5.727919999999999e-07,
+      "loss": 0.4608,
+      "step": 106850
+    },
+    {
+      "epoch": 1.069,
+      "grad_norm": 96.40602111816406,
+      "learning_rate": 5.725919999999999e-07,
+      "loss": 0.4855,
+      "step": 106900
+    },
+    {
+      "epoch": 1.0695000000000001,
+      "grad_norm": 147.89405822753906,
+      "learning_rate": 5.72392e-07,
+      "loss": 0.6196,
+      "step": 106950
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7213540077209473,
+      "learning_rate": 5.72192e-07,
+      "loss": 0.4273,
+      "step": 107000
+    },
+    {
+      "epoch": 1.0705,
+      "grad_norm": 105.5814437866211,
+      "learning_rate": 5.719919999999999e-07,
+      "loss": 0.5419,
+      "step": 107050
+    },
+    {
+      "epoch": 1.071,
+      "grad_norm": 104.46412658691406,
+      "learning_rate": 5.71792e-07,
+      "loss": 0.4807,
+      "step": 107100
+    },
+    {
+      "epoch": 1.0715,
+      "grad_norm": 51.38224792480469,
+      "learning_rate": 5.71592e-07,
+      "loss": 0.505,
+      "step": 107150
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 59.694828033447266,
+      "learning_rate": 5.713919999999999e-07,
+      "loss": 0.3088,
+      "step": 107200
+    },
+    {
+      "epoch": 1.0725,
+      "grad_norm": 17.387590408325195,
+      "learning_rate": 5.71192e-07,
+      "loss": 0.5795,
+      "step": 107250
+    },
+    {
+      "epoch": 1.073,
+      "grad_norm": 70.23807525634766,
+      "learning_rate": 5.70992e-07,
+      "loss": 0.4044,
+      "step": 107300
+    },
+    {
+      "epoch": 1.0735,
+      "grad_norm": 10.148516654968262,
+      "learning_rate": 5.707919999999999e-07,
+      "loss": 0.4085,
+      "step": 107350
+    },
+    {
+      "epoch": 1.074,
+      "grad_norm": 2.2347519397735596,
+      "learning_rate": 5.70592e-07,
+      "loss": 0.5751,
+      "step": 107400
+    },
+    {
+      "epoch": 1.0745,
+      "grad_norm": 59.01057434082031,
+      "learning_rate": 5.70392e-07,
+      "loss": 0.4502,
+      "step": 107450
+    },
+    {
+      "epoch": 1.075,
+      "grad_norm": 41.65180206298828,
+      "learning_rate": 5.701920000000001e-07,
+      "loss": 0.4107,
+      "step": 107500
+    },
+    {
+      "epoch": 1.0755,
+      "grad_norm": 63.487892150878906,
+      "learning_rate": 5.69992e-07,
+      "loss": 0.4103,
+      "step": 107550
+    },
+    {
+      "epoch": 1.076,
+      "grad_norm": 44.388309478759766,
+      "learning_rate": 5.697919999999999e-07,
+      "loss": 0.5648,
+      "step": 107600
+    },
+    {
+      "epoch": 1.0765,
+      "grad_norm": 15.944334030151367,
+      "learning_rate": 5.69592e-07,
+      "loss": 0.434,
+      "step": 107650
+    },
+    {
+      "epoch": 1.077,
+      "grad_norm": 69.580322265625,
+      "learning_rate": 5.69392e-07,
+      "loss": 0.4692,
+      "step": 107700
+    },
+    {
+      "epoch": 1.0775,
+      "grad_norm": 7.532376289367676,
+      "learning_rate": 5.69196e-07,
+      "loss": 0.486,
+      "step": 107750
+    },
+    {
+      "epoch": 1.078,
+      "grad_norm": 117.07515716552734,
+      "learning_rate": 5.68996e-07,
+      "loss": 0.5425,
+      "step": 107800
+    },
+    {
+      "epoch": 1.0785,
+      "grad_norm": 74.83606719970703,
+      "learning_rate": 5.68796e-07,
+      "loss": 0.4081,
+      "step": 107850
+    },
+    {
+      "epoch": 1.079,
+      "grad_norm": 66.38423919677734,
+      "learning_rate": 5.685959999999999e-07,
+      "loss": 0.4735,
+      "step": 107900
+    },
+    {
+      "epoch": 1.0795,
+      "grad_norm": 18.5911808013916,
+      "learning_rate": 5.68396e-07,
+      "loss": 0.5111,
+      "step": 107950
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 41.88191223144531,
+      "learning_rate": 5.68196e-07,
+      "loss": 0.4556,
+      "step": 108000
+    },
+    {
+      "epoch": 1.0805,
+      "grad_norm": 81.14200592041016,
+      "learning_rate": 5.679959999999999e-07,
+      "loss": 0.3716,
+      "step": 108050
+    },
+    {
+      "epoch": 1.081,
+      "grad_norm": 35.856998443603516,
+      "learning_rate": 5.67796e-07,
+      "loss": 0.4,
+      "step": 108100
+    },
+    {
+      "epoch": 1.0815,
+      "grad_norm": 43.836421966552734,
+      "learning_rate": 5.67596e-07,
+      "loss": 0.4874,
+      "step": 108150
+    },
+    {
+      "epoch": 1.082,
+      "grad_norm": 84.13600158691406,
+      "learning_rate": 5.67396e-07,
+      "loss": 0.4165,
+      "step": 108200
+    },
+    {
+      "epoch": 1.0825,
+      "grad_norm": 22.291542053222656,
+      "learning_rate": 5.67196e-07,
+      "loss": 0.5321,
+      "step": 108250
+    },
+    {
+      "epoch": 1.083,
+      "grad_norm": 94.12808227539062,
+      "learning_rate": 5.66996e-07,
+      "loss": 0.624,
+      "step": 108300
+    },
+    {
+      "epoch": 1.0835,
+      "grad_norm": 53.173072814941406,
+      "learning_rate": 5.667959999999999e-07,
+      "loss": 0.4123,
+      "step": 108350
+    },
+    {
+      "epoch": 1.084,
+      "grad_norm": 78.43643188476562,
+      "learning_rate": 5.66596e-07,
+      "loss": 0.5123,
+      "step": 108400
+    },
+    {
+      "epoch": 1.0845,
+      "grad_norm": 39.595787048339844,
+      "learning_rate": 5.66396e-07,
+      "loss": 0.4563,
+      "step": 108450
+    },
+    {
+      "epoch": 1.085,
+      "grad_norm": 14.890007972717285,
+      "learning_rate": 5.66196e-07,
+      "loss": 0.4253,
+      "step": 108500
+    },
+    {
+      "epoch": 1.0855,
+      "grad_norm": 91.4941635131836,
+      "learning_rate": 5.659960000000001e-07,
+      "loss": 0.3835,
+      "step": 108550
+    },
+    {
+      "epoch": 1.086,
+      "grad_norm": 74.25895690917969,
+      "learning_rate": 5.657959999999999e-07,
+      "loss": 0.4304,
+      "step": 108600
+    },
+    {
+      "epoch": 1.0865,
+      "grad_norm": 48.94237518310547,
+      "learning_rate": 5.655959999999999e-07,
+      "loss": 0.3277,
+      "step": 108650
+    },
+    {
+      "epoch": 1.087,
+      "grad_norm": 56.9554328918457,
+      "learning_rate": 5.65396e-07,
+      "loss": 0.5296,
+      "step": 108700
+    },
+    {
+      "epoch": 1.0875,
+      "grad_norm": 47.54026412963867,
+      "learning_rate": 5.65196e-07,
+      "loss": 0.3747,
+      "step": 108750
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 67.42796325683594,
+      "learning_rate": 5.64996e-07,
+      "loss": 0.4032,
+      "step": 108800
+    },
+    {
+      "epoch": 1.0885,
+      "grad_norm": 39.55683898925781,
+      "learning_rate": 5.64796e-07,
+      "loss": 0.3196,
+      "step": 108850
+    },
+    {
+      "epoch": 1.089,
+      "grad_norm": 96.23863220214844,
+      "learning_rate": 5.64596e-07,
+      "loss": 0.4986,
+      "step": 108900
+    },
+    {
+      "epoch": 1.0895,
+      "grad_norm": 6.769164085388184,
+      "learning_rate": 5.643959999999999e-07,
+      "loss": 0.4685,
+      "step": 108950
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 23.501625061035156,
+      "learning_rate": 5.64196e-07,
+      "loss": 0.6104,
+      "step": 109000
+    },
+    {
+      "epoch": 1.0905,
+      "grad_norm": 9.114831924438477,
+      "learning_rate": 5.63996e-07,
+      "loss": 0.5997,
+      "step": 109050
+    },
+    {
+      "epoch": 1.091,
+      "grad_norm": 12.518210411071777,
+      "learning_rate": 5.637959999999999e-07,
+      "loss": 0.3271,
+      "step": 109100
+    },
+    {
+      "epoch": 1.0915,
+      "grad_norm": 99.00508117675781,
+      "learning_rate": 5.63596e-07,
+      "loss": 0.5053,
+      "step": 109150
+    },
+    {
+      "epoch": 1.092,
+      "grad_norm": 10.864371299743652,
+      "learning_rate": 5.63396e-07,
+      "loss": 0.5697,
+      "step": 109200
+    },
+    {
+      "epoch": 1.0925,
+      "grad_norm": 43.38995361328125,
+      "learning_rate": 5.63196e-07,
+      "loss": 0.369,
+      "step": 109250
+    },
+    {
+      "epoch": 1.093,
+      "grad_norm": 1.3382948637008667,
+      "learning_rate": 5.62996e-07,
+      "loss": 0.5112,
+      "step": 109300
+    },
+    {
+      "epoch": 1.0935,
+      "grad_norm": 78.29591369628906,
+      "learning_rate": 5.627959999999999e-07,
+      "loss": 0.4358,
+      "step": 109350
+    },
+    {
+      "epoch": 1.094,
+      "grad_norm": 93.0485610961914,
+      "learning_rate": 5.625959999999999e-07,
+      "loss": 0.38,
+      "step": 109400
+    },
+    {
+      "epoch": 1.0945,
+      "grad_norm": 20.810575485229492,
+      "learning_rate": 5.62396e-07,
+      "loss": 0.5316,
+      "step": 109450
+    },
+    {
+      "epoch": 1.095,
+      "grad_norm": 0.7341397404670715,
+      "learning_rate": 5.62196e-07,
+      "loss": 0.3763,
+      "step": 109500
+    },
+    {
+      "epoch": 1.0955,
+      "grad_norm": 18.946903228759766,
+      "learning_rate": 5.619960000000001e-07,
+      "loss": 0.562,
+      "step": 109550
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 95.75009155273438,
+      "learning_rate": 5.61796e-07,
+      "loss": 0.4775,
+      "step": 109600
+    },
+    {
+      "epoch": 1.0965,
+      "grad_norm": 72.53479766845703,
+      "learning_rate": 5.615959999999999e-07,
+      "loss": 0.4117,
+      "step": 109650
+    },
+    {
+      "epoch": 1.097,
+      "grad_norm": 55.721282958984375,
+      "learning_rate": 5.61396e-07,
+      "loss": 0.4769,
+      "step": 109700
+    },
+    {
+      "epoch": 1.0975,
+      "grad_norm": 29.421995162963867,
+      "learning_rate": 5.61196e-07,
+      "loss": 0.4423,
+      "step": 109750
+    },
+    {
+      "epoch": 1.098,
+      "grad_norm": 62.25779342651367,
+      "learning_rate": 5.60996e-07,
+      "loss": 0.4081,
+      "step": 109800
+    },
+    {
+      "epoch": 1.0985,
+      "grad_norm": 80.15348052978516,
+      "learning_rate": 5.60796e-07,
+      "loss": 0.3587,
+      "step": 109850
+    },
+    {
+      "epoch": 1.099,
+      "grad_norm": 0.7332010865211487,
+      "learning_rate": 5.60596e-07,
+      "loss": 0.3441,
+      "step": 109900
+    },
+    {
+      "epoch": 1.0995,
+      "grad_norm": 62.10441207885742,
+      "learning_rate": 5.60396e-07,
+      "loss": 0.5722,
+      "step": 109950
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.652379035949707,
+      "learning_rate": 5.60196e-07,
+      "loss": 0.3688,
+      "step": 110000
+    },
+    {
+      "epoch": 1.1005,
+      "grad_norm": 56.87358093261719,
+      "learning_rate": 5.59996e-07,
+      "loss": 0.4241,
+      "step": 110050
+    },
+    {
+      "epoch": 1.101,
+      "grad_norm": 93.23614501953125,
+      "learning_rate": 5.597959999999999e-07,
+      "loss": 0.5574,
+      "step": 110100
+    },
+    {
+      "epoch": 1.1015,
+      "grad_norm": 34.96125793457031,
+      "learning_rate": 5.59596e-07,
+      "loss": 0.5044,
+      "step": 110150
+    },
+    {
+      "epoch": 1.102,
+      "grad_norm": 79.12371063232422,
+      "learning_rate": 5.59396e-07,
+      "loss": 0.4684,
+      "step": 110200
+    },
+    {
+      "epoch": 1.1025,
+      "grad_norm": 130.24989318847656,
+      "learning_rate": 5.59196e-07,
+      "loss": 0.4631,
+      "step": 110250
+    },
+    {
+      "epoch": 1.103,
+      "grad_norm": 121.72173309326172,
+      "learning_rate": 5.589960000000001e-07,
+      "loss": 0.4232,
+      "step": 110300
+    },
+    {
+      "epoch": 1.1035,
+      "grad_norm": 66.10565948486328,
+      "learning_rate": 5.587959999999999e-07,
+      "loss": 0.4178,
+      "step": 110350
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 70.6107406616211,
+      "learning_rate": 5.585959999999999e-07,
+      "loss": 0.5394,
+      "step": 110400
+    },
+    {
+      "epoch": 1.1045,
+      "grad_norm": 28.195026397705078,
+      "learning_rate": 5.58396e-07,
+      "loss": 0.5433,
+      "step": 110450
+    },
+    {
+      "epoch": 1.105,
+      "grad_norm": 73.64238739013672,
+      "learning_rate": 5.58196e-07,
+      "loss": 0.418,
+      "step": 110500
+    },
+    {
+      "epoch": 1.1055,
+      "grad_norm": 90.38440704345703,
+      "learning_rate": 5.57996e-07,
+      "loss": 0.6524,
+      "step": 110550
+    },
+    {
+      "epoch": 1.106,
+      "grad_norm": 7.275987148284912,
+      "learning_rate": 5.57796e-07,
+      "loss": 0.4582,
+      "step": 110600
+    },
+    {
+      "epoch": 1.1065,
+      "grad_norm": 15.238624572753906,
+      "learning_rate": 5.576e-07,
+      "loss": 0.476,
+      "step": 110650
+    },
+    {
+      "epoch": 1.107,
+      "grad_norm": 70.75611877441406,
+      "learning_rate": 5.574e-07,
+      "loss": 0.4448,
+      "step": 110700
+    },
+    {
+      "epoch": 1.1075,
+      "grad_norm": 87.86063385009766,
+      "learning_rate": 5.572e-07,
+      "loss": 0.5077,
+      "step": 110750
+    },
+    {
+      "epoch": 1.108,
+      "grad_norm": 0.4845068156719208,
+      "learning_rate": 5.57e-07,
+      "loss": 0.4965,
+      "step": 110800
+    },
+    {
+      "epoch": 1.1085,
+      "grad_norm": 2.1131792068481445,
+      "learning_rate": 5.567999999999999e-07,
+      "loss": 0.5516,
+      "step": 110850
+    },
+    {
+      "epoch": 1.109,
+      "grad_norm": 16.047956466674805,
+      "learning_rate": 5.566e-07,
+      "loss": 0.4468,
+      "step": 110900
+    },
+    {
+      "epoch": 1.1095,
+      "grad_norm": 17.67030143737793,
+      "learning_rate": 5.564e-07,
+      "loss": 0.5287,
+      "step": 110950
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 82.67327117919922,
+      "learning_rate": 5.562e-07,
+      "loss": 0.3612,
+      "step": 111000
+    },
+    {
+      "epoch": 1.1105,
+      "grad_norm": 1.270891785621643,
+      "learning_rate": 5.560000000000001e-07,
+      "loss": 0.396,
+      "step": 111050
+    },
+    {
+      "epoch": 1.111,
+      "grad_norm": 75.45249938964844,
+      "learning_rate": 5.557999999999999e-07,
+      "loss": 0.3984,
+      "step": 111100
+    },
+    {
+      "epoch": 1.1115,
+      "grad_norm": 19.672801971435547,
+      "learning_rate": 5.555999999999999e-07,
+      "loss": 0.5507,
+      "step": 111150
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 6.31002140045166,
+      "learning_rate": 5.554e-07,
+      "loss": 0.3859,
+      "step": 111200
+    },
+    {
+      "epoch": 1.1125,
+      "grad_norm": 85.23663330078125,
+      "learning_rate": 5.552e-07,
+      "loss": 0.4728,
+      "step": 111250
+    },
+    {
+      "epoch": 1.113,
+      "grad_norm": 88.14482879638672,
+      "learning_rate": 5.55e-07,
+      "loss": 0.5598,
+      "step": 111300
+    },
+    {
+      "epoch": 1.1135,
+      "grad_norm": 94.62075805664062,
+      "learning_rate": 5.548e-07,
+      "loss": 0.4012,
+      "step": 111350
+    },
+    {
+      "epoch": 1.114,
+      "grad_norm": 51.954891204833984,
+      "learning_rate": 5.546e-07,
+      "loss": 0.3856,
+      "step": 111400
+    },
+    {
+      "epoch": 1.1145,
+      "grad_norm": 75.34689331054688,
+      "learning_rate": 5.543999999999999e-07,
+      "loss": 0.3746,
+      "step": 111450
+    },
+    {
+      "epoch": 1.115,
+      "grad_norm": 72.60891723632812,
+      "learning_rate": 5.542e-07,
+      "loss": 0.5083,
+      "step": 111500
+    },
+    {
+      "epoch": 1.1155,
+      "grad_norm": 39.89189147949219,
+      "learning_rate": 5.54e-07,
+      "loss": 0.4479,
+      "step": 111550
+    },
+    {
+      "epoch": 1.116,
+      "grad_norm": 4.4385833740234375,
+      "learning_rate": 5.537999999999999e-07,
+      "loss": 0.3594,
+      "step": 111600
+    },
+    {
+      "epoch": 1.1165,
+      "grad_norm": 0.6604699492454529,
+      "learning_rate": 5.536e-07,
+      "loss": 0.3674,
+      "step": 111650
+    },
+    {
+      "epoch": 1.117,
+      "grad_norm": 49.04147720336914,
+      "learning_rate": 5.534e-07,
+      "loss": 0.4568,
+      "step": 111700
+    },
+    {
+      "epoch": 1.1175,
+      "grad_norm": 42.06256866455078,
+      "learning_rate": 5.532e-07,
+      "loss": 0.5162,
+      "step": 111750
+    },
+    {
+      "epoch": 1.1179999999999999,
+      "grad_norm": 102.51911163330078,
+      "learning_rate": 5.53e-07,
+      "loss": 0.4049,
+      "step": 111800
+    },
+    {
+      "epoch": 1.1185,
+      "grad_norm": 44.5997428894043,
+      "learning_rate": 5.527999999999999e-07,
+      "loss": 0.4188,
+      "step": 111850
+    },
+    {
+      "epoch": 1.119,
+      "grad_norm": 25.700366973876953,
+      "learning_rate": 5.525999999999999e-07,
+      "loss": 0.3842,
+      "step": 111900
+    },
+    {
+      "epoch": 1.1195,
+      "grad_norm": 16.742462158203125,
+      "learning_rate": 5.524e-07,
+      "loss": 0.5652,
+      "step": 111950
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 52.463157653808594,
+      "learning_rate": 5.522e-07,
+      "loss": 0.4582,
+      "step": 112000
+    },
+    {
+      "epoch": 1.1205,
+      "grad_norm": 37.043521881103516,
+      "learning_rate": 5.520000000000001e-07,
+      "loss": 0.4157,
+      "step": 112050
+    },
+    {
+      "epoch": 1.121,
+      "grad_norm": 1.2325105667114258,
+      "learning_rate": 5.518e-07,
+      "loss": 0.5246,
+      "step": 112100
+    },
+    {
+      "epoch": 1.1215,
+      "grad_norm": 21.83551597595215,
+      "learning_rate": 5.515999999999999e-07,
+      "loss": 0.4453,
+      "step": 112150
+    },
+    {
+      "epoch": 1.1219999999999999,
+      "grad_norm": 3.408167839050293,
+      "learning_rate": 5.514e-07,
+      "loss": 0.4424,
+      "step": 112200
+    },
+    {
+      "epoch": 1.1225,
+      "grad_norm": 86.52948760986328,
+      "learning_rate": 5.512e-07,
+      "loss": 0.516,
+      "step": 112250
+    },
+    {
+      "epoch": 1.123,
+      "grad_norm": 2.4071269035339355,
+      "learning_rate": 5.51e-07,
+      "loss": 0.304,
+      "step": 112300
+    },
+    {
+      "epoch": 1.1235,
+      "grad_norm": 74.5964126586914,
+      "learning_rate": 5.508e-07,
+      "loss": 0.2992,
+      "step": 112350
+    },
+    {
+      "epoch": 1.124,
+      "grad_norm": 52.07612991333008,
+      "learning_rate": 5.506e-07,
+      "loss": 0.4152,
+      "step": 112400
+    },
+    {
+      "epoch": 1.1245,
+      "grad_norm": 66.46135711669922,
+      "learning_rate": 5.504e-07,
+      "loss": 0.4892,
+      "step": 112450
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 19.547279357910156,
+      "learning_rate": 5.502e-07,
+      "loss": 0.3387,
+      "step": 112500
+    },
+    {
+      "epoch": 1.1255,
+      "grad_norm": 44.79277420043945,
+      "learning_rate": 5.5e-07,
+      "loss": 0.459,
+      "step": 112550
+    },
+    {
+      "epoch": 1.126,
+      "grad_norm": 112.7216567993164,
+      "learning_rate": 5.497999999999999e-07,
+      "loss": 0.3898,
+      "step": 112600
+    },
+    {
+      "epoch": 1.1265,
+      "grad_norm": 14.867293357849121,
+      "learning_rate": 5.496e-07,
+      "loss": 0.5735,
+      "step": 112650
+    },
+    {
+      "epoch": 1.127,
+      "grad_norm": 98.45852661132812,
+      "learning_rate": 5.494e-07,
+      "loss": 0.5208,
+      "step": 112700
+    },
+    {
+      "epoch": 1.1275,
+      "grad_norm": 27.679582595825195,
+      "learning_rate": 5.492e-07,
+      "loss": 0.429,
+      "step": 112750
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.8725939989089966,
+      "learning_rate": 5.490000000000001e-07,
+      "loss": 0.4026,
+      "step": 112800
+    },
+    {
+      "epoch": 1.1285,
+      "grad_norm": 9.201322555541992,
+      "learning_rate": 5.487999999999999e-07,
+      "loss": 0.5991,
+      "step": 112850
+    },
+    {
+      "epoch": 1.129,
+      "grad_norm": 42.13934326171875,
+      "learning_rate": 5.485999999999999e-07,
+      "loss": 0.4351,
+      "step": 112900
+    },
+    {
+      "epoch": 1.1295,
+      "grad_norm": 22.487741470336914,
+      "learning_rate": 5.484e-07,
+      "loss": 0.3695,
+      "step": 112950
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 78.89250183105469,
+      "learning_rate": 5.482e-07,
+      "loss": 0.6346,
+      "step": 113000
+    },
+    {
+      "epoch": 1.1305,
+      "grad_norm": 123.7279052734375,
+      "learning_rate": 5.48e-07,
+      "loss": 0.5355,
+      "step": 113050
+    },
+    {
+      "epoch": 1.131,
+      "grad_norm": 19.976831436157227,
+      "learning_rate": 5.478e-07,
+      "loss": 0.3888,
+      "step": 113100
+    },
+    {
+      "epoch": 1.1315,
+      "grad_norm": 36.473289489746094,
+      "learning_rate": 5.476e-07,
+      "loss": 0.6061,
+      "step": 113150
+    },
+    {
+      "epoch": 1.1320000000000001,
+      "grad_norm": 33.37117004394531,
+      "learning_rate": 5.473999999999999e-07,
+      "loss": 0.3772,
+      "step": 113200
+    },
+    {
+      "epoch": 1.1325,
+      "grad_norm": 45.76275634765625,
+      "learning_rate": 5.472e-07,
+      "loss": 0.4798,
+      "step": 113250
+    },
+    {
+      "epoch": 1.133,
+      "grad_norm": 67.44389343261719,
+      "learning_rate": 5.47e-07,
+      "loss": 0.5103,
+      "step": 113300
+    },
+    {
+      "epoch": 1.1335,
+      "grad_norm": 77.15213775634766,
+      "learning_rate": 5.467999999999999e-07,
+      "loss": 0.4921,
+      "step": 113350
+    },
+    {
+      "epoch": 1.134,
+      "grad_norm": 109.74034881591797,
+      "learning_rate": 5.466e-07,
+      "loss": 0.5541,
+      "step": 113400
+    },
+    {
+      "epoch": 1.1345,
+      "grad_norm": 49.160316467285156,
+      "learning_rate": 5.464e-07,
+      "loss": 0.4071,
+      "step": 113450
+    },
+    {
+      "epoch": 1.135,
+      "grad_norm": 0.6659963130950928,
+      "learning_rate": 5.462e-07,
+      "loss": 0.4504,
+      "step": 113500
+    },
+    {
+      "epoch": 1.1355,
+      "grad_norm": 129.61778259277344,
+      "learning_rate": 5.46e-07,
+      "loss": 0.4083,
+      "step": 113550
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 51.32670211791992,
+      "learning_rate": 5.457999999999999e-07,
+      "loss": 0.4323,
+      "step": 113600
+    },
+    {
+      "epoch": 1.1365,
+      "grad_norm": 94.61491394042969,
+      "learning_rate": 5.455999999999999e-07,
+      "loss": 0.4553,
+      "step": 113650
+    },
+    {
+      "epoch": 1.137,
+      "grad_norm": 94.95795440673828,
+      "learning_rate": 5.454e-07,
+      "loss": 0.3422,
+      "step": 113700
+    },
+    {
+      "epoch": 1.1375,
+      "grad_norm": 45.39287185668945,
+      "learning_rate": 5.452e-07,
+      "loss": 0.4915,
+      "step": 113750
+    },
+    {
+      "epoch": 1.138,
+      "grad_norm": 29.25931739807129,
+      "learning_rate": 5.45e-07,
+      "loss": 0.3838,
+      "step": 113800
+    },
+    {
+      "epoch": 1.1385,
+      "grad_norm": 60.96989822387695,
+      "learning_rate": 5.448e-07,
+      "loss": 0.5589,
+      "step": 113850
+    },
+    {
+      "epoch": 1.139,
+      "grad_norm": 45.03044128417969,
+      "learning_rate": 5.445999999999999e-07,
+      "loss": 0.4151,
+      "step": 113900
+    },
+    {
+      "epoch": 1.1395,
+      "grad_norm": 6.3732404708862305,
+      "learning_rate": 5.443999999999999e-07,
+      "loss": 0.3358,
+      "step": 113950
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 18.30240249633789,
+      "learning_rate": 5.442e-07,
+      "loss": 0.5567,
+      "step": 114000
+    },
+    {
+      "epoch": 1.1405,
+      "grad_norm": 47.34935760498047,
+      "learning_rate": 5.44e-07,
+      "loss": 0.4934,
+      "step": 114050
+    },
+    {
+      "epoch": 1.141,
+      "grad_norm": 12.485174179077148,
+      "learning_rate": 5.437999999999999e-07,
+      "loss": 0.4781,
+      "step": 114100
+    },
+    {
+      "epoch": 1.1415,
+      "grad_norm": 6.659138202667236,
+      "learning_rate": 5.436e-07,
+      "loss": 0.4272,
+      "step": 114150
+    },
+    {
+      "epoch": 1.142,
+      "grad_norm": 17.311927795410156,
+      "learning_rate": 5.434e-07,
+      "loss": 0.3993,
+      "step": 114200
+    },
+    {
+      "epoch": 1.1425,
+      "grad_norm": 70.42869567871094,
+      "learning_rate": 5.431999999999999e-07,
+      "loss": 0.4323,
+      "step": 114250
+    },
+    {
+      "epoch": 1.143,
+      "grad_norm": 44.21244812011719,
+      "learning_rate": 5.43e-07,
+      "loss": 0.4421,
+      "step": 114300
+    },
+    {
+      "epoch": 1.1435,
+      "grad_norm": 66.3972396850586,
+      "learning_rate": 5.427999999999999e-07,
+      "loss": 0.4448,
+      "step": 114350
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 31.286418914794922,
+      "learning_rate": 5.425999999999999e-07,
+      "loss": 0.5305,
+      "step": 114400
+    },
+    {
+      "epoch": 1.1445,
+      "grad_norm": 79.53046417236328,
+      "learning_rate": 5.424e-07,
+      "loss": 0.5406,
+      "step": 114450
+    },
+    {
+      "epoch": 1.145,
+      "grad_norm": 1.268337607383728,
+      "learning_rate": 5.422e-07,
+      "loss": 0.3917,
+      "step": 114500
+    },
+    {
+      "epoch": 1.1455,
+      "grad_norm": 27.05938148498535,
+      "learning_rate": 5.420000000000001e-07,
+      "loss": 0.5285,
+      "step": 114550
+    },
+    {
+      "epoch": 1.146,
+      "grad_norm": 24.848316192626953,
+      "learning_rate": 5.417999999999999e-07,
+      "loss": 0.5525,
+      "step": 114600
+    },
+    {
+      "epoch": 1.1465,
+      "grad_norm": 21.377756118774414,
+      "learning_rate": 5.415999999999999e-07,
+      "loss": 0.5307,
+      "step": 114650
+    },
+    {
+      "epoch": 1.147,
+      "grad_norm": 3.2962486743927,
+      "learning_rate": 5.414e-07,
+      "loss": 0.4409,
+      "step": 114700
+    },
+    {
+      "epoch": 1.1475,
+      "grad_norm": 77.59668731689453,
+      "learning_rate": 5.41204e-07,
+      "loss": 0.5064,
+      "step": 114750
+    },
+    {
+      "epoch": 1.148,
+      "grad_norm": 82.37496185302734,
+      "learning_rate": 5.41004e-07,
+      "loss": 0.3953,
+      "step": 114800
+    },
+    {
+      "epoch": 1.1485,
+      "grad_norm": 85.80392456054688,
+      "learning_rate": 5.40804e-07,
+      "loss": 0.5122,
+      "step": 114850
+    },
+    {
+      "epoch": 1.149,
+      "grad_norm": 81.59890747070312,
+      "learning_rate": 5.40604e-07,
+      "loss": 0.5849,
+      "step": 114900
+    },
+    {
+      "epoch": 1.1495,
+      "grad_norm": 58.08201217651367,
+      "learning_rate": 5.404039999999999e-07,
+      "loss": 0.4393,
+      "step": 114950
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 78.7620849609375,
+      "learning_rate": 5.40204e-07,
+      "loss": 0.4161,
+      "step": 115000
+    },
+    {
+      "epoch": 1.1505,
+      "grad_norm": 84.27385711669922,
+      "learning_rate": 5.40004e-07,
+      "loss": 0.3664,
+      "step": 115050
+    },
+    {
+      "epoch": 1.151,
+      "grad_norm": 62.63862609863281,
+      "learning_rate": 5.398039999999999e-07,
+      "loss": 0.4925,
+      "step": 115100
+    },
+    {
+      "epoch": 1.1515,
+      "grad_norm": 67.16043090820312,
+      "learning_rate": 5.39604e-07,
+      "loss": 0.5412,
+      "step": 115150
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 53.16383361816406,
+      "learning_rate": 5.39404e-07,
+      "loss": 0.404,
+      "step": 115200
+    },
+    {
+      "epoch": 1.1525,
+      "grad_norm": 39.47855758666992,
+      "learning_rate": 5.39204e-07,
+      "loss": 0.4474,
+      "step": 115250
+    },
+    {
+      "epoch": 1.153,
+      "grad_norm": 1.1669477224349976,
+      "learning_rate": 5.390040000000001e-07,
+      "loss": 0.4366,
+      "step": 115300
+    },
+    {
+      "epoch": 1.1535,
+      "grad_norm": 67.96946716308594,
+      "learning_rate": 5.388039999999999e-07,
+      "loss": 0.478,
+      "step": 115350
+    },
+    {
+      "epoch": 1.154,
+      "grad_norm": 26.770727157592773,
+      "learning_rate": 5.386039999999999e-07,
+      "loss": 0.3817,
+      "step": 115400
+    },
+    {
+      "epoch": 1.1545,
+      "grad_norm": 57.50968933105469,
+      "learning_rate": 5.38404e-07,
+      "loss": 0.3984,
+      "step": 115450
+    },
+    {
+      "epoch": 1.155,
+      "grad_norm": 33.6568603515625,
+      "learning_rate": 5.38204e-07,
+      "loss": 0.4302,
+      "step": 115500
+    },
+    {
+      "epoch": 1.1555,
+      "grad_norm": 1.8125121593475342,
+      "learning_rate": 5.38004e-07,
+      "loss": 0.4788,
+      "step": 115550
+    },
+    {
+      "epoch": 1.156,
+      "grad_norm": 57.99422073364258,
+      "learning_rate": 5.37804e-07,
+      "loss": 0.2844,
+      "step": 115600
+    },
+    {
+      "epoch": 1.1565,
+      "grad_norm": 63.1536865234375,
+      "learning_rate": 5.37608e-07,
+      "loss": 0.3902,
+      "step": 115650
+    },
+    {
+      "epoch": 1.157,
+      "grad_norm": 15.688610076904297,
+      "learning_rate": 5.374079999999999e-07,
+      "loss": 0.3493,
+      "step": 115700
+    },
+    {
+      "epoch": 1.1575,
+      "grad_norm": 77.63614654541016,
+      "learning_rate": 5.37208e-07,
+      "loss": 0.5133,
+      "step": 115750
+    },
+    {
+      "epoch": 1.158,
+      "grad_norm": 36.73105239868164,
+      "learning_rate": 5.37008e-07,
+      "loss": 0.4251,
+      "step": 115800
+    },
+    {
+      "epoch": 1.1585,
+      "grad_norm": 42.460872650146484,
+      "learning_rate": 5.368079999999999e-07,
+      "loss": 0.5139,
+      "step": 115850
+    },
+    {
+      "epoch": 1.159,
+      "grad_norm": 73.16018676757812,
+      "learning_rate": 5.36608e-07,
+      "loss": 0.3794,
+      "step": 115900
+    },
+    {
+      "epoch": 1.1595,
+      "grad_norm": 42.55072021484375,
+      "learning_rate": 5.36408e-07,
+      "loss": 0.3831,
+      "step": 115950
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 32.332515716552734,
+      "learning_rate": 5.36208e-07,
+      "loss": 0.3343,
+      "step": 116000
+    },
+    {
+      "epoch": 1.1605,
+      "grad_norm": 47.2687873840332,
+      "learning_rate": 5.36008e-07,
+      "loss": 0.5091,
+      "step": 116050
+    },
+    {
+      "epoch": 1.161,
+      "grad_norm": 95.39241790771484,
+      "learning_rate": 5.358079999999999e-07,
+      "loss": 0.4826,
+      "step": 116100
+    },
+    {
+      "epoch": 1.1615,
+      "grad_norm": 83.02808380126953,
+      "learning_rate": 5.356079999999999e-07,
+      "loss": 0.4376,
+      "step": 116150
+    },
+    {
+      "epoch": 1.162,
+      "grad_norm": 59.88267517089844,
+      "learning_rate": 5.35408e-07,
+      "loss": 0.4086,
+      "step": 116200
+    },
+    {
+      "epoch": 1.1625,
+      "grad_norm": 62.45161056518555,
+      "learning_rate": 5.35208e-07,
+      "loss": 0.4202,
+      "step": 116250
+    },
+    {
+      "epoch": 1.163,
+      "grad_norm": 117.84101104736328,
+      "learning_rate": 5.35008e-07,
+      "loss": 0.464,
+      "step": 116300
+    },
+    {
+      "epoch": 1.1635,
+      "grad_norm": 65.64312744140625,
+      "learning_rate": 5.34808e-07,
+      "loss": 0.5111,
+      "step": 116350
+    },
+    {
+      "epoch": 1.164,
+      "grad_norm": 45.327972412109375,
+      "learning_rate": 5.346079999999999e-07,
+      "loss": 0.4727,
+      "step": 116400
+    },
+    {
+      "epoch": 1.1645,
+      "grad_norm": 61.47356414794922,
+      "learning_rate": 5.344079999999999e-07,
+      "loss": 0.3288,
+      "step": 116450
+    },
+    {
+      "epoch": 1.165,
+      "grad_norm": 34.35345458984375,
+      "learning_rate": 5.34208e-07,
+      "loss": 0.365,
+      "step": 116500
+    },
+    {
+      "epoch": 1.1655,
+      "grad_norm": 5.033834934234619,
+      "learning_rate": 5.34008e-07,
+      "loss": 0.4617,
+      "step": 116550
+    },
+    {
+      "epoch": 1.166,
+      "grad_norm": 0.281574010848999,
+      "learning_rate": 5.338079999999999e-07,
+      "loss": 0.4622,
+      "step": 116600
+    },
+    {
+      "epoch": 1.1665,
+      "grad_norm": 51.69894027709961,
+      "learning_rate": 5.33608e-07,
+      "loss": 0.4423,
+      "step": 116650
+    },
+    {
+      "epoch": 1.167,
+      "grad_norm": 36.404632568359375,
+      "learning_rate": 5.33408e-07,
+      "loss": 0.5443,
+      "step": 116700
+    },
+    {
+      "epoch": 1.1675,
+      "grad_norm": 78.08043670654297,
+      "learning_rate": 5.332079999999999e-07,
+      "loss": 0.4774,
+      "step": 116750
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 30.812881469726562,
+      "learning_rate": 5.33008e-07,
+      "loss": 0.5483,
+      "step": 116800
+    },
+    {
+      "epoch": 1.1685,
+      "grad_norm": 0.9321984648704529,
+      "learning_rate": 5.328079999999999e-07,
+      "loss": 0.3477,
+      "step": 116850
+    },
+    {
+      "epoch": 1.169,
+      "grad_norm": 72.7139892578125,
+      "learning_rate": 5.326079999999999e-07,
+      "loss": 0.5847,
+      "step": 116900
+    },
+    {
+      "epoch": 1.1695,
+      "grad_norm": 50.76666259765625,
+      "learning_rate": 5.32408e-07,
+      "loss": 0.3907,
+      "step": 116950
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 96.60762786865234,
+      "learning_rate": 5.32208e-07,
+      "loss": 0.672,
+      "step": 117000
+    },
+    {
+      "epoch": 1.1705,
+      "grad_norm": 65.07868957519531,
+      "learning_rate": 5.320080000000001e-07,
+      "loss": 0.4654,
+      "step": 117050
+    },
+    {
+      "epoch": 1.171,
+      "grad_norm": 41.71099090576172,
+      "learning_rate": 5.318079999999999e-07,
+      "loss": 0.3896,
+      "step": 117100
+    },
+    {
+      "epoch": 1.1715,
+      "grad_norm": 30.449907302856445,
+      "learning_rate": 5.316079999999999e-07,
+      "loss": 0.5155,
+      "step": 117150
+    },
+    {
+      "epoch": 1.172,
+      "grad_norm": 103.2293701171875,
+      "learning_rate": 5.31408e-07,
+      "loss": 0.3301,
+      "step": 117200
+    },
+    {
+      "epoch": 1.1724999999999999,
+      "grad_norm": 54.75368881225586,
+      "learning_rate": 5.31208e-07,
+      "loss": 0.4718,
+      "step": 117250
+    },
+    {
+      "epoch": 1.173,
+      "grad_norm": 75.55032348632812,
+      "learning_rate": 5.31008e-07,
+      "loss": 0.4543,
+      "step": 117300
+    },
+    {
+      "epoch": 1.1735,
+      "grad_norm": 54.67997741699219,
+      "learning_rate": 5.30808e-07,
+      "loss": 0.4419,
+      "step": 117350
+    },
+    {
+      "epoch": 1.174,
+      "grad_norm": 91.2918472290039,
+      "learning_rate": 5.30608e-07,
+      "loss": 0.4074,
+      "step": 117400
+    },
+    {
+      "epoch": 1.1745,
+      "grad_norm": 73.0489501953125,
+      "learning_rate": 5.304079999999999e-07,
+      "loss": 0.4729,
+      "step": 117450
+    },
+    {
+      "epoch": 1.175,
+      "grad_norm": 3.4424524307250977,
+      "learning_rate": 5.30208e-07,
+      "loss": 0.4709,
+      "step": 117500
+    },
+    {
+      "epoch": 1.1755,
+      "grad_norm": 64.21658325195312,
+      "learning_rate": 5.30008e-07,
+      "loss": 0.4925,
+      "step": 117550
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 24.437042236328125,
+      "learning_rate": 5.298079999999999e-07,
+      "loss": 0.4695,
+      "step": 117600
+    },
+    {
+      "epoch": 1.1764999999999999,
+      "grad_norm": 85.16154479980469,
+      "learning_rate": 5.29608e-07,
+      "loss": 0.4352,
+      "step": 117650
+    },
+    {
+      "epoch": 1.177,
+      "grad_norm": 34.37142562866211,
+      "learning_rate": 5.29408e-07,
+      "loss": 0.4595,
+      "step": 117700
+    },
+    {
+      "epoch": 1.1775,
+      "grad_norm": 103.63298034667969,
+      "learning_rate": 5.29208e-07,
+      "loss": 0.4311,
+      "step": 117750
+    },
+    {
+      "epoch": 1.178,
+      "grad_norm": 72.38675689697266,
+      "learning_rate": 5.29008e-07,
+      "loss": 0.4235,
+      "step": 117800
+    },
+    {
+      "epoch": 1.1785,
+      "grad_norm": 4.414365291595459,
+      "learning_rate": 5.288079999999999e-07,
+      "loss": 0.462,
+      "step": 117850
+    },
+    {
+      "epoch": 1.179,
+      "grad_norm": 74.01151275634766,
+      "learning_rate": 5.286079999999999e-07,
+      "loss": 0.5448,
+      "step": 117900
+    },
+    {
+      "epoch": 1.1795,
+      "grad_norm": 88.3164291381836,
+      "learning_rate": 5.28408e-07,
+      "loss": 0.2708,
+      "step": 117950
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 75.72642517089844,
+      "learning_rate": 5.28208e-07,
+      "loss": 0.3894,
+      "step": 118000
+    },
+    {
+      "epoch": 1.1804999999999999,
+      "grad_norm": 63.216453552246094,
+      "learning_rate": 5.28008e-07,
+      "loss": 0.4749,
+      "step": 118050
+    },
+    {
+      "epoch": 1.181,
+      "grad_norm": 39.14889144897461,
+      "learning_rate": 5.278080000000001e-07,
+      "loss": 0.3814,
+      "step": 118100
+    },
+    {
+      "epoch": 1.1815,
+      "grad_norm": 56.19570541381836,
+      "learning_rate": 5.276079999999999e-07,
+      "loss": 0.349,
+      "step": 118150
+    },
+    {
+      "epoch": 1.182,
+      "grad_norm": 99.52997589111328,
+      "learning_rate": 5.274079999999999e-07,
+      "loss": 0.6551,
+      "step": 118200
+    },
+    {
+      "epoch": 1.1825,
+      "grad_norm": 27.029600143432617,
+      "learning_rate": 5.27208e-07,
+      "loss": 0.5198,
+      "step": 118250
+    },
+    {
+      "epoch": 1.183,
+      "grad_norm": 33.4561882019043,
+      "learning_rate": 5.27008e-07,
+      "loss": 0.4874,
+      "step": 118300
+    },
+    {
+      "epoch": 1.1835,
+      "grad_norm": 24.337610244750977,
+      "learning_rate": 5.26808e-07,
+      "loss": 0.5505,
+      "step": 118350
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 39.1830940246582,
+      "learning_rate": 5.26608e-07,
+      "loss": 0.4899,
+      "step": 118400
+    },
+    {
+      "epoch": 1.1844999999999999,
+      "grad_norm": 32.634098052978516,
+      "learning_rate": 5.26408e-07,
+      "loss": 0.354,
+      "step": 118450
+    },
+    {
+      "epoch": 1.185,
+      "grad_norm": 102.42017364501953,
+      "learning_rate": 5.262079999999999e-07,
+      "loss": 0.3999,
+      "step": 118500
+    },
+    {
+      "epoch": 1.1855,
+      "grad_norm": 1.6400700807571411,
+      "learning_rate": 5.26008e-07,
+      "loss": 0.4722,
+      "step": 118550
+    },
+    {
+      "epoch": 1.186,
+      "grad_norm": 56.43583297729492,
+      "learning_rate": 5.25808e-07,
+      "loss": 0.5107,
+      "step": 118600
+    },
+    {
+      "epoch": 1.1865,
+      "grad_norm": 4.176393508911133,
+      "learning_rate": 5.256079999999999e-07,
+      "loss": 0.5332,
+      "step": 118650
+    },
+    {
+      "epoch": 1.187,
+      "grad_norm": 74.65667724609375,
+      "learning_rate": 5.25408e-07,
+      "loss": 0.347,
+      "step": 118700
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 61.597930908203125,
+      "learning_rate": 5.25208e-07,
+      "loss": 0.5037,
+      "step": 118750
+    },
+    {
+      "epoch": 1.188,
+      "grad_norm": 20.81285285949707,
+      "learning_rate": 5.25008e-07,
+      "loss": 0.4662,
+      "step": 118800
+    },
+    {
+      "epoch": 1.1885,
+      "grad_norm": 16.560958862304688,
+      "learning_rate": 5.24808e-07,
+      "loss": 0.4709,
+      "step": 118850
+    },
+    {
+      "epoch": 1.189,
+      "grad_norm": 85.1524429321289,
+      "learning_rate": 5.246079999999999e-07,
+      "loss": 0.3761,
+      "step": 118900
+    },
+    {
+      "epoch": 1.1895,
+      "grad_norm": 38.864070892333984,
+      "learning_rate": 5.244079999999999e-07,
+      "loss": 0.3661,
+      "step": 118950
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 25.94007682800293,
+      "learning_rate": 5.24208e-07,
+      "loss": 0.3894,
+      "step": 119000
+    },
+    {
+      "epoch": 1.1905000000000001,
+      "grad_norm": 89.40354919433594,
+      "learning_rate": 5.24008e-07,
+      "loss": 0.6054,
+      "step": 119050
+    },
+    {
+      "epoch": 1.191,
+      "grad_norm": 28.081167221069336,
+      "learning_rate": 5.238080000000001e-07,
+      "loss": 0.4478,
+      "step": 119100
+    },
+    {
+      "epoch": 1.1915,
+      "grad_norm": 86.49946594238281,
+      "learning_rate": 5.23608e-07,
+      "loss": 0.4845,
+      "step": 119150
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 29.06024169921875,
+      "learning_rate": 5.234079999999999e-07,
+      "loss": 0.3214,
+      "step": 119200
+    },
+    {
+      "epoch": 1.1925,
+      "grad_norm": 21.522964477539062,
+      "learning_rate": 5.23208e-07,
+      "loss": 0.3218,
+      "step": 119250
+    },
+    {
+      "epoch": 1.193,
+      "grad_norm": 77.7065200805664,
+      "learning_rate": 5.23008e-07,
+      "loss": 0.422,
+      "step": 119300
+    },
+    {
+      "epoch": 1.1935,
+      "grad_norm": 6.1106343269348145,
+      "learning_rate": 5.22808e-07,
+      "loss": 0.5586,
+      "step": 119350
+    },
+    {
+      "epoch": 1.194,
+      "grad_norm": 9.23585319519043,
+      "learning_rate": 5.22608e-07,
+      "loss": 0.4294,
+      "step": 119400
+    },
+    {
+      "epoch": 1.1945000000000001,
+      "grad_norm": 79.62310791015625,
+      "learning_rate": 5.22408e-07,
+      "loss": 0.4129,
+      "step": 119450
+    },
+    {
+      "epoch": 1.195,
+      "grad_norm": 18.00088882446289,
+      "learning_rate": 5.22208e-07,
+      "loss": 0.5814,
+      "step": 119500
+    },
+    {
+      "epoch": 1.1955,
+      "grad_norm": 110.30831146240234,
+      "learning_rate": 5.22008e-07,
+      "loss": 0.5494,
+      "step": 119550
+    },
+    {
+      "epoch": 1.196,
+      "grad_norm": 53.79940414428711,
+      "learning_rate": 5.21808e-07,
+      "loss": 0.3932,
+      "step": 119600
+    },
+    {
+      "epoch": 1.1965,
+      "grad_norm": 76.86608123779297,
+      "learning_rate": 5.216079999999999e-07,
+      "loss": 0.5098,
+      "step": 119650
+    },
+    {
+      "epoch": 1.197,
+      "grad_norm": 17.632843017578125,
+      "learning_rate": 5.21408e-07,
+      "loss": 0.5619,
+      "step": 119700
+    },
+    {
+      "epoch": 1.1975,
+      "grad_norm": 121.02203369140625,
+      "learning_rate": 5.21208e-07,
+      "loss": 0.478,
+      "step": 119750
+    },
+    {
+      "epoch": 1.198,
+      "grad_norm": 7.7555084228515625,
+      "learning_rate": 5.21008e-07,
+      "loss": 0.3932,
+      "step": 119800
+    },
+    {
+      "epoch": 1.1985000000000001,
+      "grad_norm": 79.47029113769531,
+      "learning_rate": 5.208080000000001e-07,
+      "loss": 0.4755,
+      "step": 119850
+    },
+    {
+      "epoch": 1.199,
+      "grad_norm": 85.48977661132812,
+      "learning_rate": 5.206079999999999e-07,
+      "loss": 0.5045,
+      "step": 119900
+    },
+    {
+      "epoch": 1.1995,
+      "grad_norm": 27.96647834777832,
+      "learning_rate": 5.204079999999999e-07,
+      "loss": 0.4199,
+      "step": 119950
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 93.02487182617188,
+      "learning_rate": 5.20208e-07,
+      "loss": 0.471,
+      "step": 120000
+    },
+    {
+      "epoch": 1.2005,
+      "grad_norm": 27.727758407592773,
+      "learning_rate": 5.20008e-07,
+      "loss": 0.4306,
+      "step": 120050
+    },
+    {
+      "epoch": 1.201,
+      "grad_norm": 82.77517700195312,
+      "learning_rate": 5.19808e-07,
+      "loss": 0.4908,
+      "step": 120100
+    },
+    {
+      "epoch": 1.2015,
+      "grad_norm": 83.77193450927734,
+      "learning_rate": 5.19608e-07,
+      "loss": 0.415,
+      "step": 120150
+    },
+    {
+      "epoch": 1.202,
+      "grad_norm": 1.6518832445144653,
+      "learning_rate": 5.19408e-07,
+      "loss": 0.6042,
+      "step": 120200
+    },
+    {
+      "epoch": 1.2025000000000001,
+      "grad_norm": 39.49478530883789,
+      "learning_rate": 5.192079999999999e-07,
+      "loss": 0.4554,
+      "step": 120250
+    },
+    {
+      "epoch": 1.203,
+      "grad_norm": 62.075286865234375,
+      "learning_rate": 5.19008e-07,
+      "loss": 0.4344,
+      "step": 120300
+    },
+    {
+      "epoch": 1.2035,
+      "grad_norm": 0.7993803024291992,
+      "learning_rate": 5.18808e-07,
+      "loss": 0.3913,
+      "step": 120350
+    },
+    {
+      "epoch": 1.204,
+      "grad_norm": 0.20502905547618866,
+      "learning_rate": 5.186079999999999e-07,
+      "loss": 0.2958,
+      "step": 120400
+    },
+    {
+      "epoch": 1.2045,
+      "grad_norm": 23.898794174194336,
+      "learning_rate": 5.18412e-07,
+      "loss": 0.4188,
+      "step": 120450
+    },
+    {
+      "epoch": 1.205,
+      "grad_norm": 4.550911903381348,
+      "learning_rate": 5.18212e-07,
+      "loss": 0.4426,
+      "step": 120500
+    },
+    {
+      "epoch": 1.2055,
+      "grad_norm": 39.06218719482422,
+      "learning_rate": 5.18012e-07,
+      "loss": 0.4951,
+      "step": 120550
+    },
+    {
+      "epoch": 1.206,
+      "grad_norm": 77.20858001708984,
+      "learning_rate": 5.178120000000001e-07,
+      "loss": 0.4668,
+      "step": 120600
+    },
+    {
+      "epoch": 1.2065,
+      "grad_norm": 27.37639617919922,
+      "learning_rate": 5.176119999999999e-07,
+      "loss": 0.5083,
+      "step": 120650
+    },
+    {
+      "epoch": 1.207,
+      "grad_norm": 102.03683471679688,
+      "learning_rate": 5.174119999999999e-07,
+      "loss": 0.4479,
+      "step": 120700
+    },
+    {
+      "epoch": 1.2075,
+      "grad_norm": 10.521049499511719,
+      "learning_rate": 5.17212e-07,
+      "loss": 0.4924,
+      "step": 120750
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 50.4854621887207,
+      "learning_rate": 5.17012e-07,
+      "loss": 0.595,
+      "step": 120800
+    },
+    {
+      "epoch": 1.2085,
+      "grad_norm": 60.808013916015625,
+      "learning_rate": 5.16812e-07,
+      "loss": 0.3966,
+      "step": 120850
+    },
+    {
+      "epoch": 1.209,
+      "grad_norm": 35.996952056884766,
+      "learning_rate": 5.16612e-07,
+      "loss": 0.3948,
+      "step": 120900
+    },
+    {
+      "epoch": 1.2095,
+      "grad_norm": 83.63671112060547,
+      "learning_rate": 5.16412e-07,
+      "loss": 0.6021,
+      "step": 120950
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 92.66004180908203,
+      "learning_rate": 5.162119999999999e-07,
+      "loss": 0.3842,
+      "step": 121000
+    },
+    {
+      "epoch": 1.2105,
+      "grad_norm": 13.310367584228516,
+      "learning_rate": 5.16012e-07,
+      "loss": 0.6077,
+      "step": 121050
+    },
+    {
+      "epoch": 1.211,
+      "grad_norm": 4.458984375,
+      "learning_rate": 5.15812e-07,
+      "loss": 0.3837,
+      "step": 121100
+    },
+    {
+      "epoch": 1.2115,
+      "grad_norm": 39.17095184326172,
+      "learning_rate": 5.156119999999999e-07,
+      "loss": 0.4516,
+      "step": 121150
+    },
+    {
+      "epoch": 1.212,
+      "grad_norm": 77.5641860961914,
+      "learning_rate": 5.15412e-07,
+      "loss": 0.4531,
+      "step": 121200
+    },
+    {
+      "epoch": 1.2125,
+      "grad_norm": 31.501514434814453,
+      "learning_rate": 5.15212e-07,
+      "loss": 0.4466,
+      "step": 121250
+    },
+    {
+      "epoch": 1.213,
+      "grad_norm": 5.8556413650512695,
+      "learning_rate": 5.15012e-07,
+      "loss": 0.473,
+      "step": 121300
+    },
+    {
+      "epoch": 1.2135,
+      "grad_norm": 37.5422477722168,
+      "learning_rate": 5.14812e-07,
+      "loss": 0.4125,
+      "step": 121350
+    },
+    {
+      "epoch": 1.214,
+      "grad_norm": 2.340611457824707,
+      "learning_rate": 5.146119999999999e-07,
+      "loss": 0.4659,
+      "step": 121400
+    },
+    {
+      "epoch": 1.2145,
+      "grad_norm": 16.7429141998291,
+      "learning_rate": 5.144119999999999e-07,
+      "loss": 0.4964,
+      "step": 121450
+    },
+    {
+      "epoch": 1.215,
+      "grad_norm": 122.42613220214844,
+      "learning_rate": 5.14212e-07,
+      "loss": 0.403,
+      "step": 121500
+    },
+    {
+      "epoch": 1.2155,
+      "grad_norm": 89.16505432128906,
+      "learning_rate": 5.14012e-07,
+      "loss": 0.474,
+      "step": 121550
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 41.55615997314453,
+      "learning_rate": 5.138120000000001e-07,
+      "loss": 0.4665,
+      "step": 121600
+    },
+    {
+      "epoch": 1.2165,
+      "grad_norm": 4.50972843170166,
+      "learning_rate": 5.13612e-07,
+      "loss": 0.4449,
+      "step": 121650
+    },
+    {
+      "epoch": 1.217,
+      "grad_norm": 45.18336486816406,
+      "learning_rate": 5.134119999999999e-07,
+      "loss": 0.4837,
+      "step": 121700
+    },
+    {
+      "epoch": 1.2175,
+      "grad_norm": 15.911096572875977,
+      "learning_rate": 5.13212e-07,
+      "loss": 0.4563,
+      "step": 121750
+    },
+    {
+      "epoch": 1.218,
+      "grad_norm": 32.39420700073242,
+      "learning_rate": 5.13012e-07,
+      "loss": 0.3398,
+      "step": 121800
+    },
+    {
+      "epoch": 1.2185,
+      "grad_norm": 67.41966247558594,
+      "learning_rate": 5.12812e-07,
+      "loss": 0.4374,
+      "step": 121850
+    },
+    {
+      "epoch": 1.219,
+      "grad_norm": 5.196559429168701,
+      "learning_rate": 5.12612e-07,
+      "loss": 0.5112,
+      "step": 121900
+    },
+    {
+      "epoch": 1.2195,
+      "grad_norm": 50.85737991333008,
+      "learning_rate": 5.12412e-07,
+      "loss": 0.5262,
+      "step": 121950
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6516265869140625,
+      "learning_rate": 5.12212e-07,
+      "loss": 0.3481,
+      "step": 122000
+    },
+    {
+      "epoch": 1.2205,
+      "grad_norm": 39.09980392456055,
+      "learning_rate": 5.12012e-07,
+      "loss": 0.3903,
+      "step": 122050
+    },
+    {
+      "epoch": 1.221,
+      "grad_norm": 45.83763122558594,
+      "learning_rate": 5.11812e-07,
+      "loss": 0.572,
+      "step": 122100
+    },
+    {
+      "epoch": 1.2215,
+      "grad_norm": 78.31942749023438,
+      "learning_rate": 5.116119999999999e-07,
+      "loss": 0.4883,
+      "step": 122150
+    },
+    {
+      "epoch": 1.222,
+      "grad_norm": 64.6844253540039,
+      "learning_rate": 5.11412e-07,
+      "loss": 0.5336,
+      "step": 122200
+    },
+    {
+      "epoch": 1.2225,
+      "grad_norm": 1.514631986618042,
+      "learning_rate": 5.11212e-07,
+      "loss": 0.4101,
+      "step": 122250
+    },
+    {
+      "epoch": 1.223,
+      "grad_norm": 47.222328186035156,
+      "learning_rate": 5.11012e-07,
+      "loss": 0.5162,
+      "step": 122300
+    },
+    {
+      "epoch": 1.2235,
+      "grad_norm": 54.371952056884766,
+      "learning_rate": 5.108120000000001e-07,
+      "loss": 0.5621,
+      "step": 122350
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 62.68058395385742,
+      "learning_rate": 5.106119999999999e-07,
+      "loss": 0.4995,
+      "step": 122400
+    },
+    {
+      "epoch": 1.2245,
+      "grad_norm": 39.96584701538086,
+      "learning_rate": 5.104119999999999e-07,
+      "loss": 0.3692,
+      "step": 122450
+    },
+    {
+      "epoch": 1.225,
+      "grad_norm": 3.750762462615967,
+      "learning_rate": 5.10212e-07,
+      "loss": 0.4632,
+      "step": 122500
+    },
+    {
+      "epoch": 1.2255,
+      "grad_norm": 5.8588080406188965,
+      "learning_rate": 5.10012e-07,
+      "loss": 0.4385,
+      "step": 122550
+    },
+    {
+      "epoch": 1.226,
+      "grad_norm": 27.323911666870117,
+      "learning_rate": 5.09812e-07,
+      "loss": 0.3817,
+      "step": 122600
+    },
+    {
+      "epoch": 1.2265,
+      "grad_norm": 78.1037826538086,
+      "learning_rate": 5.09612e-07,
+      "loss": 0.5332,
+      "step": 122650
+    },
+    {
+      "epoch": 1.227,
+      "grad_norm": 30.815229415893555,
+      "learning_rate": 5.09412e-07,
+      "loss": 0.517,
+      "step": 122700
+    },
+    {
+      "epoch": 1.2275,
+      "grad_norm": 8.353984832763672,
+      "learning_rate": 5.092119999999999e-07,
+      "loss": 0.4976,
+      "step": 122750
+    },
+    {
+      "epoch": 1.228,
+      "grad_norm": 119.27523040771484,
+      "learning_rate": 5.09012e-07,
+      "loss": 0.4711,
+      "step": 122800
+    },
+    {
+      "epoch": 1.2285,
+      "grad_norm": 73.34131622314453,
+      "learning_rate": 5.08812e-07,
+      "loss": 0.5385,
+      "step": 122850
+    },
+    {
+      "epoch": 1.229,
+      "grad_norm": 0.49966078996658325,
+      "learning_rate": 5.086159999999999e-07,
+      "loss": 0.4272,
+      "step": 122900
+    },
+    {
+      "epoch": 1.2295,
+      "grad_norm": 60.721370697021484,
+      "learning_rate": 5.0842e-07,
+      "loss": 0.4648,
+      "step": 122950
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 87.89384460449219,
+      "learning_rate": 5.0822e-07,
+      "loss": 0.4228,
+      "step": 123000
+    },
+    {
+      "epoch": 1.2305,
+      "grad_norm": 79.43232727050781,
+      "learning_rate": 5.0802e-07,
+      "loss": 0.4643,
+      "step": 123050
+    },
+    {
+      "epoch": 1.231,
+      "grad_norm": 22.63487434387207,
+      "learning_rate": 5.078200000000001e-07,
+      "loss": 0.5683,
+      "step": 123100
+    },
+    {
+      "epoch": 1.2315,
+      "grad_norm": 79.83702850341797,
+      "learning_rate": 5.076199999999999e-07,
+      "loss": 0.5672,
+      "step": 123150
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 56.245948791503906,
+      "learning_rate": 5.074199999999999e-07,
+      "loss": 0.4281,
+      "step": 123200
+    },
+    {
+      "epoch": 1.2325,
+      "grad_norm": 17.55223846435547,
+      "learning_rate": 5.0722e-07,
+      "loss": 0.4716,
+      "step": 123250
+    },
+    {
+      "epoch": 1.233,
+      "grad_norm": 10.06242561340332,
+      "learning_rate": 5.0702e-07,
+      "loss": 0.4489,
+      "step": 123300
+    },
+    {
+      "epoch": 1.2335,
+      "grad_norm": 11.893889427185059,
+      "learning_rate": 5.0682e-07,
+      "loss": 0.4702,
+      "step": 123350
+    },
+    {
+      "epoch": 1.234,
+      "grad_norm": 105.41064453125,
+      "learning_rate": 5.0662e-07,
+      "loss": 0.4019,
+      "step": 123400
+    },
+    {
+      "epoch": 1.2345,
+      "grad_norm": 86.82225799560547,
+      "learning_rate": 5.0642e-07,
+      "loss": 0.4129,
+      "step": 123450
+    },
+    {
+      "epoch": 1.2349999999999999,
+      "grad_norm": 68.03509521484375,
+      "learning_rate": 5.062199999999999e-07,
+      "loss": 0.5526,
+      "step": 123500
+    },
+    {
+      "epoch": 1.2355,
+      "grad_norm": 27.85662841796875,
+      "learning_rate": 5.0602e-07,
+      "loss": 0.388,
+      "step": 123550
+    },
+    {
+      "epoch": 1.236,
+      "grad_norm": 15.7693452835083,
+      "learning_rate": 5.0582e-07,
+      "loss": 0.6257,
+      "step": 123600
+    },
+    {
+      "epoch": 1.2365,
+      "grad_norm": 8.797830581665039,
+      "learning_rate": 5.056199999999999e-07,
+      "loss": 0.3529,
+      "step": 123650
+    },
+    {
+      "epoch": 1.237,
+      "grad_norm": 46.14826202392578,
+      "learning_rate": 5.0542e-07,
+      "loss": 0.4523,
+      "step": 123700
+    },
+    {
+      "epoch": 1.2375,
+      "grad_norm": 16.78753662109375,
+      "learning_rate": 5.0522e-07,
+      "loss": 0.4304,
+      "step": 123750
+    },
+    {
+      "epoch": 1.238,
+      "grad_norm": 90.13909912109375,
+      "learning_rate": 5.0502e-07,
+      "loss": 0.4219,
+      "step": 123800
+    },
+    {
+      "epoch": 1.2385,
+      "grad_norm": 20.657495498657227,
+      "learning_rate": 5.0482e-07,
+      "loss": 0.4275,
+      "step": 123850
+    },
+    {
+      "epoch": 1.2389999999999999,
+      "grad_norm": 117.53194427490234,
+      "learning_rate": 5.046199999999999e-07,
+      "loss": 0.5234,
+      "step": 123900
+    },
+    {
+      "epoch": 1.2395,
+      "grad_norm": 44.009521484375,
+      "learning_rate": 5.044199999999999e-07,
+      "loss": 0.3952,
+      "step": 123950
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 62.8895263671875,
+      "learning_rate": 5.0422e-07,
+      "loss": 0.3973,
+      "step": 124000
+    },
+    {
+      "epoch": 1.2405,
+      "grad_norm": 83.4673080444336,
+      "learning_rate": 5.0402e-07,
+      "loss": 0.4622,
+      "step": 124050
+    },
+    {
+      "epoch": 1.241,
+      "grad_norm": 69.27863311767578,
+      "learning_rate": 5.038200000000001e-07,
+      "loss": 0.3788,
+      "step": 124100
+    },
+    {
+      "epoch": 1.2415,
+      "grad_norm": 27.948728561401367,
+      "learning_rate": 5.0362e-07,
+      "loss": 0.4593,
+      "step": 124150
+    },
+    {
+      "epoch": 1.242,
+      "grad_norm": 17.01951789855957,
+      "learning_rate": 5.034199999999999e-07,
+      "loss": 0.4347,
+      "step": 124200
+    },
+    {
+      "epoch": 1.2425,
+      "grad_norm": 95.70691680908203,
+      "learning_rate": 5.0322e-07,
+      "loss": 0.4042,
+      "step": 124250
+    },
+    {
+      "epoch": 1.2429999999999999,
+      "grad_norm": 120.20599365234375,
+      "learning_rate": 5.0302e-07,
+      "loss": 0.4595,
+      "step": 124300
+    },
+    {
+      "epoch": 1.2435,
+      "grad_norm": 119.31159973144531,
+      "learning_rate": 5.0282e-07,
+      "loss": 0.4708,
+      "step": 124350
+    },
+    {
+      "epoch": 1.244,
+      "grad_norm": 72.69075775146484,
+      "learning_rate": 5.0262e-07,
+      "loss": 0.5223,
+      "step": 124400
+    },
+    {
+      "epoch": 1.2445,
+      "grad_norm": 65.67170715332031,
+      "learning_rate": 5.0242e-07,
+      "loss": 0.5593,
+      "step": 124450
+    },
+    {
+      "epoch": 1.245,
+      "grad_norm": 32.64443588256836,
+      "learning_rate": 5.0222e-07,
+      "loss": 0.532,
+      "step": 124500
+    },
+    {
+      "epoch": 1.2455,
+      "grad_norm": 6.584482192993164,
+      "learning_rate": 5.0202e-07,
+      "loss": 0.4046,
+      "step": 124550
+    },
+    {
+      "epoch": 1.246,
+      "grad_norm": 26.74454116821289,
+      "learning_rate": 5.0182e-07,
+      "loss": 0.493,
+      "step": 124600
+    },
+    {
+      "epoch": 1.2465,
+      "grad_norm": 3.494159460067749,
+      "learning_rate": 5.016199999999999e-07,
+      "loss": 0.395,
+      "step": 124650
+    },
+    {
+      "epoch": 1.2469999999999999,
+      "grad_norm": 12.93493938446045,
+      "learning_rate": 5.0142e-07,
+      "loss": 0.3823,
+      "step": 124700
+    },
+    {
+      "epoch": 1.2475,
+      "grad_norm": 58.11399459838867,
+      "learning_rate": 5.0122e-07,
+      "loss": 0.4471,
+      "step": 124750
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 51.41734313964844,
+      "learning_rate": 5.0102e-07,
+      "loss": 0.4074,
+      "step": 124800
+    },
+    {
+      "epoch": 1.2485,
+      "grad_norm": 76.81995391845703,
+      "learning_rate": 5.008200000000001e-07,
+      "loss": 0.4224,
+      "step": 124850
+    },
+    {
+      "epoch": 1.249,
+      "grad_norm": 48.812599182128906,
+      "learning_rate": 5.006199999999999e-07,
+      "loss": 0.4964,
+      "step": 124900
+    },
+    {
+      "epoch": 1.2495,
+      "grad_norm": 62.64362716674805,
+      "learning_rate": 5.004199999999999e-07,
+      "loss": 0.423,
+      "step": 124950
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 88.53785705566406,
+      "learning_rate": 5.0022e-07,
+      "loss": 0.4533,
+      "step": 125000
+    },
+    {
+      "epoch": 1.2505,
+      "grad_norm": 88.8245620727539,
+      "learning_rate": 5.0002e-07,
+      "loss": 0.5211,
+      "step": 125050
+    },
+    {
+      "epoch": 1.251,
+      "grad_norm": 43.50419616699219,
+      "learning_rate": 4.99824e-07,
+      "loss": 0.4007,
+      "step": 125100
+    },
+    {
+      "epoch": 1.2515,
+      "grad_norm": 11.175210952758789,
+      "learning_rate": 4.99624e-07,
+      "loss": 0.5217,
+      "step": 125150
+    },
+    {
+      "epoch": 1.252,
+      "grad_norm": 50.50872039794922,
+      "learning_rate": 4.99424e-07,
+      "loss": 0.4395,
+      "step": 125200
+    },
+    {
+      "epoch": 1.2525,
+      "grad_norm": 28.890705108642578,
+      "learning_rate": 4.992239999999999e-07,
+      "loss": 0.3575,
+      "step": 125250
+    },
+    {
+      "epoch": 1.2530000000000001,
+      "grad_norm": 2.7780516147613525,
+      "learning_rate": 4.99024e-07,
+      "loss": 0.4679,
+      "step": 125300
+    },
+    {
+      "epoch": 1.2535,
+      "grad_norm": 74.0434341430664,
+      "learning_rate": 4.988239999999999e-07,
+      "loss": 0.4339,
+      "step": 125350
+    },
+    {
+      "epoch": 1.254,
+      "grad_norm": 1.51035737991333,
+      "learning_rate": 4.98624e-07,
+      "loss": 0.4149,
+      "step": 125400
+    },
+    {
+      "epoch": 1.2545,
+      "grad_norm": 11.029925346374512,
+      "learning_rate": 4.98424e-07,
+      "loss": 0.4752,
+      "step": 125450
+    },
+    {
+      "epoch": 1.255,
+      "grad_norm": 4.947063446044922,
+      "learning_rate": 4.98224e-07,
+      "loss": 0.5536,
+      "step": 125500
+    },
+    {
+      "epoch": 1.2555,
+      "grad_norm": 1.3308128118515015,
+      "learning_rate": 4.98024e-07,
+      "loss": 0.3657,
+      "step": 125550
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 61.55348205566406,
+      "learning_rate": 4.978239999999999e-07,
+      "loss": 0.5046,
+      "step": 125600
+    },
+    {
+      "epoch": 1.2565,
+      "grad_norm": 128.6259765625,
+      "learning_rate": 4.97624e-07,
+      "loss": 0.4097,
+      "step": 125650
+    },
+    {
+      "epoch": 1.2570000000000001,
+      "grad_norm": 7.346100807189941,
+      "learning_rate": 4.974239999999999e-07,
+      "loss": 0.4103,
+      "step": 125700
+    },
+    {
+      "epoch": 1.2575,
+      "grad_norm": 8.92379093170166,
+      "learning_rate": 4.97224e-07,
+      "loss": 0.2637,
+      "step": 125750
+    },
+    {
+      "epoch": 1.258,
+      "grad_norm": 42.0355110168457,
+      "learning_rate": 4.97024e-07,
+      "loss": 0.3719,
+      "step": 125800
+    },
+    {
+      "epoch": 1.2585,
+      "grad_norm": 77.65760803222656,
+      "learning_rate": 4.96824e-07,
+      "loss": 0.3427,
+      "step": 125850
+    },
+    {
+      "epoch": 1.259,
+      "grad_norm": 95.41388702392578,
+      "learning_rate": 4.96624e-07,
+      "loss": 0.4427,
+      "step": 125900
+    },
+    {
+      "epoch": 1.2595,
+      "grad_norm": 38.658050537109375,
+      "learning_rate": 4.964239999999999e-07,
+      "loss": 0.5894,
+      "step": 125950
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.92863130569458,
+      "learning_rate": 4.962239999999999e-07,
+      "loss": 0.5318,
+      "step": 126000
+    },
+    {
+      "epoch": 1.2605,
+      "grad_norm": 132.43011474609375,
+      "learning_rate": 4.96024e-07,
+      "loss": 0.4031,
+      "step": 126050
+    },
+    {
+      "epoch": 1.2610000000000001,
+      "grad_norm": 51.750030517578125,
+      "learning_rate": 4.95824e-07,
+      "loss": 0.4338,
+      "step": 126100
+    },
+    {
+      "epoch": 1.2615,
+      "grad_norm": 4.630794525146484,
+      "learning_rate": 4.95624e-07,
+      "loss": 0.4925,
+      "step": 126150
+    },
+    {
+      "epoch": 1.262,
+      "grad_norm": 91.11925506591797,
+      "learning_rate": 4.95424e-07,
+      "loss": 0.4702,
+      "step": 126200
+    },
+    {
+      "epoch": 1.2625,
+      "grad_norm": 68.52376556396484,
+      "learning_rate": 4.95224e-07,
+      "loss": 0.6044,
+      "step": 126250
+    },
+    {
+      "epoch": 1.263,
+      "grad_norm": 84.54537200927734,
+      "learning_rate": 4.950239999999999e-07,
+      "loss": 0.5117,
+      "step": 126300
+    },
+    {
+      "epoch": 1.2635,
+      "grad_norm": 6.372779846191406,
+      "learning_rate": 4.948239999999999e-07,
+      "loss": 0.3654,
+      "step": 126350
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 10.201987266540527,
+      "learning_rate": 4.94624e-07,
+      "loss": 0.5223,
+      "step": 126400
+    },
+    {
+      "epoch": 1.2645,
+      "grad_norm": 48.930599212646484,
+      "learning_rate": 4.944239999999999e-07,
+      "loss": 0.4524,
+      "step": 126450
+    },
+    {
+      "epoch": 1.2650000000000001,
+      "grad_norm": 63.114070892333984,
+      "learning_rate": 4.94224e-07,
+      "loss": 0.5026,
+      "step": 126500
+    },
+    {
+      "epoch": 1.2655,
+      "grad_norm": 64.93250274658203,
+      "learning_rate": 4.94024e-07,
+      "loss": 0.4449,
+      "step": 126550
+    },
+    {
+      "epoch": 1.266,
+      "grad_norm": 18.260643005371094,
+      "learning_rate": 4.93824e-07,
+      "loss": 0.4546,
+      "step": 126600
+    },
+    {
+      "epoch": 1.2665,
+      "grad_norm": 56.187164306640625,
+      "learning_rate": 4.93624e-07,
+      "loss": 0.4692,
+      "step": 126650
+    },
+    {
+      "epoch": 1.267,
+      "grad_norm": 75.86105346679688,
+      "learning_rate": 4.934239999999999e-07,
+      "loss": 0.4319,
+      "step": 126700
+    },
+    {
+      "epoch": 1.2675,
+      "grad_norm": 59.311798095703125,
+      "learning_rate": 4.93228e-07,
+      "loss": 0.4846,
+      "step": 126750
+    },
+    {
+      "epoch": 1.268,
+      "grad_norm": 34.68928146362305,
+      "learning_rate": 4.93028e-07,
+      "loss": 0.3693,
+      "step": 126800
+    },
+    {
+      "epoch": 1.2685,
+      "grad_norm": 58.48672866821289,
+      "learning_rate": 4.92828e-07,
+      "loss": 0.5163,
+      "step": 126850
+    },
+    {
+      "epoch": 1.2690000000000001,
+      "grad_norm": 6.793260097503662,
+      "learning_rate": 4.92628e-07,
+      "loss": 0.4033,
+      "step": 126900
+    },
+    {
+      "epoch": 1.2695,
+      "grad_norm": 108.26202392578125,
+      "learning_rate": 4.92428e-07,
+      "loss": 0.4662,
+      "step": 126950
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 62.59315872192383,
+      "learning_rate": 4.92228e-07,
+      "loss": 0.427,
+      "step": 127000
+    },
+    {
+      "epoch": 1.2705,
+      "grad_norm": 7.085406303405762,
+      "learning_rate": 4.92028e-07,
+      "loss": 0.356,
+      "step": 127050
+    },
+    {
+      "epoch": 1.271,
+      "grad_norm": 27.684703826904297,
+      "learning_rate": 4.918279999999999e-07,
+      "loss": 0.3432,
+      "step": 127100
+    },
+    {
+      "epoch": 1.2715,
+      "grad_norm": 89.28202819824219,
+      "learning_rate": 4.91628e-07,
+      "loss": 0.4642,
+      "step": 127150
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 81.8940658569336,
+      "learning_rate": 4.91428e-07,
+      "loss": 0.4793,
+      "step": 127200
+    },
+    {
+      "epoch": 1.2725,
+      "grad_norm": 57.01283645629883,
+      "learning_rate": 4.91228e-07,
+      "loss": 0.6866,
+      "step": 127250
+    },
+    {
+      "epoch": 1.2730000000000001,
+      "grad_norm": 7.828011989593506,
+      "learning_rate": 4.91028e-07,
+      "loss": 0.4175,
+      "step": 127300
+    },
+    {
+      "epoch": 1.2735,
+      "grad_norm": 6.988823890686035,
+      "learning_rate": 4.90828e-07,
+      "loss": 0.4843,
+      "step": 127350
+    },
+    {
+      "epoch": 1.274,
+      "grad_norm": 129.013916015625,
+      "learning_rate": 4.906279999999999e-07,
+      "loss": 0.4135,
+      "step": 127400
+    },
+    {
+      "epoch": 1.2745,
+      "grad_norm": 2.377220392227173,
+      "learning_rate": 4.904279999999999e-07,
+      "loss": 0.4675,
+      "step": 127450
+    },
+    {
+      "epoch": 1.275,
+      "grad_norm": 73.6028060913086,
+      "learning_rate": 4.90228e-07,
+      "loss": 0.4821,
+      "step": 127500
+    },
+    {
+      "epoch": 1.2755,
+      "grad_norm": 53.05596160888672,
+      "learning_rate": 4.90028e-07,
+      "loss": 0.3646,
+      "step": 127550
+    },
+    {
+      "epoch": 1.276,
+      "grad_norm": 2.0764973163604736,
+      "learning_rate": 4.89828e-07,
+      "loss": 0.4229,
+      "step": 127600
+    },
+    {
+      "epoch": 1.2765,
+      "grad_norm": 36.13945770263672,
+      "learning_rate": 4.89628e-07,
+      "loss": 0.4127,
+      "step": 127650
+    },
+    {
+      "epoch": 1.2770000000000001,
+      "grad_norm": 66.94934844970703,
+      "learning_rate": 4.89428e-07,
+      "loss": 0.4685,
+      "step": 127700
+    },
+    {
+      "epoch": 1.2775,
+      "grad_norm": 100.14254760742188,
+      "learning_rate": 4.892279999999999e-07,
+      "loss": 0.5389,
+      "step": 127750
+    },
+    {
+      "epoch": 1.278,
+      "grad_norm": 56.245418548583984,
+      "learning_rate": 4.89028e-07,
+      "loss": 0.4782,
+      "step": 127800
+    },
+    {
+      "epoch": 1.2785,
+      "grad_norm": 2.368191957473755,
+      "learning_rate": 4.88828e-07,
+      "loss": 0.3856,
+      "step": 127850
+    },
+    {
+      "epoch": 1.279,
+      "grad_norm": 125.4609375,
+      "learning_rate": 4.88628e-07,
+      "loss": 0.5209,
+      "step": 127900
+    },
+    {
+      "epoch": 1.2795,
+      "grad_norm": 9.785893440246582,
+      "learning_rate": 4.88428e-07,
+      "loss": 0.3902,
+      "step": 127950
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 76.34190368652344,
+      "learning_rate": 4.88228e-07,
+      "loss": 0.4521,
+      "step": 128000
+    },
+    {
+      "epoch": 1.2805,
+      "grad_norm": 37.4355354309082,
+      "learning_rate": 4.88028e-07,
+      "loss": 0.3555,
+      "step": 128050
+    },
+    {
+      "epoch": 1.2810000000000001,
+      "grad_norm": 9.940116882324219,
+      "learning_rate": 4.878279999999999e-07,
+      "loss": 0.3704,
+      "step": 128100
+    },
+    {
+      "epoch": 1.2814999999999999,
+      "grad_norm": 6.346107006072998,
+      "learning_rate": 4.87628e-07,
+      "loss": 0.3613,
+      "step": 128150
+    },
+    {
+      "epoch": 1.282,
+      "grad_norm": 20.553491592407227,
+      "learning_rate": 4.874279999999999e-07,
+      "loss": 0.4186,
+      "step": 128200
+    },
+    {
+      "epoch": 1.2825,
+      "grad_norm": 31.738750457763672,
+      "learning_rate": 4.87228e-07,
+      "loss": 0.5275,
+      "step": 128250
+    },
+    {
+      "epoch": 1.283,
+      "grad_norm": 32.1572151184082,
+      "learning_rate": 4.87028e-07,
+      "loss": 0.5016,
+      "step": 128300
+    },
+    {
+      "epoch": 1.2835,
+      "grad_norm": 60.900787353515625,
+      "learning_rate": 4.86828e-07,
+      "loss": 0.5527,
+      "step": 128350
+    },
+    {
+      "epoch": 1.284,
+      "grad_norm": 37.76549530029297,
+      "learning_rate": 4.86628e-07,
+      "loss": 0.3499,
+      "step": 128400
+    },
+    {
+      "epoch": 1.2845,
+      "grad_norm": 16.48982810974121,
+      "learning_rate": 4.86428e-07,
+      "loss": 0.3264,
+      "step": 128450
+    },
+    {
+      "epoch": 1.285,
+      "grad_norm": 11.211465835571289,
+      "learning_rate": 4.862279999999999e-07,
+      "loss": 0.3776,
+      "step": 128500
+    },
+    {
+      "epoch": 1.2854999999999999,
+      "grad_norm": 4.8228583335876465,
+      "learning_rate": 4.86028e-07,
+      "loss": 0.497,
+      "step": 128550
+    },
+    {
+      "epoch": 1.286,
+      "grad_norm": 51.52220916748047,
+      "learning_rate": 4.85828e-07,
+      "loss": 0.4283,
+      "step": 128600
+    },
+    {
+      "epoch": 1.2865,
+      "grad_norm": 116.35945892333984,
+      "learning_rate": 4.85628e-07,
+      "loss": 0.4819,
+      "step": 128650
+    },
+    {
+      "epoch": 1.287,
+      "grad_norm": 16.744569778442383,
+      "learning_rate": 4.85428e-07,
+      "loss": 0.4903,
+      "step": 128700
+    },
+    {
+      "epoch": 1.2875,
+      "grad_norm": 83.41133880615234,
+      "learning_rate": 4.85228e-07,
+      "loss": 0.4459,
+      "step": 128750
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 14.692625999450684,
+      "learning_rate": 4.850279999999999e-07,
+      "loss": 0.3043,
+      "step": 128800
+    },
+    {
+      "epoch": 1.2885,
+      "grad_norm": 56.206844329833984,
+      "learning_rate": 4.848279999999999e-07,
+      "loss": 0.4554,
+      "step": 128850
+    },
+    {
+      "epoch": 1.289,
+      "grad_norm": 89.15912628173828,
+      "learning_rate": 4.84628e-07,
+      "loss": 0.5114,
+      "step": 128900
+    },
+    {
+      "epoch": 1.2894999999999999,
+      "grad_norm": 101.62841033935547,
+      "learning_rate": 4.84428e-07,
+      "loss": 0.491,
+      "step": 128950
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 65.63429260253906,
+      "learning_rate": 4.84228e-07,
+      "loss": 0.5268,
+      "step": 129000
+    },
+    {
+      "epoch": 1.2905,
+      "grad_norm": 50.63749694824219,
+      "learning_rate": 4.84028e-07,
+      "loss": 0.4863,
+      "step": 129050
+    },
+    {
+      "epoch": 1.291,
+      "grad_norm": 9.616476058959961,
+      "learning_rate": 4.83828e-07,
+      "loss": 0.3873,
+      "step": 129100
+    },
+    {
+      "epoch": 1.2915,
+      "grad_norm": 38.35820388793945,
+      "learning_rate": 4.836279999999999e-07,
+      "loss": 0.2707,
+      "step": 129150
+    },
+    {
+      "epoch": 1.292,
+      "grad_norm": 7.243841171264648,
+      "learning_rate": 4.83428e-07,
+      "loss": 0.3993,
+      "step": 129200
+    },
+    {
+      "epoch": 1.2925,
+      "grad_norm": 4.215170860290527,
+      "learning_rate": 4.83228e-07,
+      "loss": 0.3448,
+      "step": 129250
+    },
+    {
+      "epoch": 1.293,
+      "grad_norm": 34.597991943359375,
+      "learning_rate": 4.83028e-07,
+      "loss": 0.3508,
+      "step": 129300
+    },
+    {
+      "epoch": 1.2934999999999999,
+      "grad_norm": 65.05717468261719,
+      "learning_rate": 4.82828e-07,
+      "loss": 0.3527,
+      "step": 129350
+    },
+    {
+      "epoch": 1.294,
+      "grad_norm": 13.01252269744873,
+      "learning_rate": 4.82628e-07,
+      "loss": 0.4451,
+      "step": 129400
+    },
+    {
+      "epoch": 1.2945,
+      "grad_norm": 13.404302597045898,
+      "learning_rate": 4.82428e-07,
+      "loss": 0.3919,
+      "step": 129450
+    },
+    {
+      "epoch": 1.295,
+      "grad_norm": 62.99480056762695,
+      "learning_rate": 4.822279999999999e-07,
+      "loss": 0.4649,
+      "step": 129500
+    },
+    {
+      "epoch": 1.2955,
+      "grad_norm": 45.989017486572266,
+      "learning_rate": 4.82028e-07,
+      "loss": 0.479,
+      "step": 129550
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 34.542762756347656,
+      "learning_rate": 4.818279999999999e-07,
+      "loss": 0.4426,
+      "step": 129600
+    },
+    {
+      "epoch": 1.2965,
+      "grad_norm": 41.29684066772461,
+      "learning_rate": 4.81628e-07,
+      "loss": 0.4677,
+      "step": 129650
+    },
+    {
+      "epoch": 1.297,
+      "grad_norm": 92.34574127197266,
+      "learning_rate": 4.81428e-07,
+      "loss": 0.3883,
+      "step": 129700
+    },
+    {
+      "epoch": 1.2974999999999999,
+      "grad_norm": 27.71408462524414,
+      "learning_rate": 4.81228e-07,
+      "loss": 0.4247,
+      "step": 129750
+    },
+    {
+      "epoch": 1.298,
+      "grad_norm": 6.5036492347717285,
+      "learning_rate": 4.81028e-07,
+      "loss": 0.4296,
+      "step": 129800
+    },
+    {
+      "epoch": 1.2985,
+      "grad_norm": 17.653621673583984,
+      "learning_rate": 4.808279999999999e-07,
+      "loss": 0.5943,
+      "step": 129850
+    },
+    {
+      "epoch": 1.299,
+      "grad_norm": 17.503652572631836,
+      "learning_rate": 4.806319999999999e-07,
+      "loss": 0.4773,
+      "step": 129900
+    },
+    {
+      "epoch": 1.2995,
+      "grad_norm": 34.05784225463867,
+      "learning_rate": 4.80432e-07,
+      "loss": 0.455,
+      "step": 129950
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.221923351287842,
+      "learning_rate": 4.80232e-07,
+      "loss": 0.3534,
+      "step": 130000
+    },
+    {
+      "epoch": 1.3005,
+      "grad_norm": 44.165382385253906,
+      "learning_rate": 4.80032e-07,
+      "loss": 0.4089,
+      "step": 130050
+    },
+    {
+      "epoch": 1.301,
+      "grad_norm": 19.523441314697266,
+      "learning_rate": 4.79832e-07,
+      "loss": 0.5113,
+      "step": 130100
+    },
+    {
+      "epoch": 1.3014999999999999,
+      "grad_norm": 15.177743911743164,
+      "learning_rate": 4.79632e-07,
+      "loss": 0.4199,
+      "step": 130150
+    },
+    {
+      "epoch": 1.302,
+      "grad_norm": 0.24970543384552002,
+      "learning_rate": 4.794320000000001e-07,
+      "loss": 0.2723,
+      "step": 130200
+    },
+    {
+      "epoch": 1.3025,
+      "grad_norm": 80.9247817993164,
+      "learning_rate": 4.792319999999999e-07,
+      "loss": 0.4628,
+      "step": 130250
+    },
+    {
+      "epoch": 1.303,
+      "grad_norm": 24.861129760742188,
+      "learning_rate": 4.79032e-07,
+      "loss": 0.4166,
+      "step": 130300
+    },
+    {
+      "epoch": 1.3035,
+      "grad_norm": 4.55272912979126,
+      "learning_rate": 4.78832e-07,
+      "loss": 0.4463,
+      "step": 130350
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 11.789888381958008,
+      "learning_rate": 4.78632e-07,
+      "loss": 0.5848,
+      "step": 130400
+    },
+    {
+      "epoch": 1.3045,
+      "grad_norm": 3.2823047637939453,
+      "learning_rate": 4.78432e-07,
+      "loss": 0.4772,
+      "step": 130450
+    },
+    {
+      "epoch": 1.305,
+      "grad_norm": 8.338168144226074,
+      "learning_rate": 4.78232e-07,
+      "loss": 0.3896,
+      "step": 130500
+    },
+    {
+      "epoch": 1.3054999999999999,
+      "grad_norm": 84.38629150390625,
+      "learning_rate": 4.78032e-07,
+      "loss": 0.4665,
+      "step": 130550
+    },
+    {
+      "epoch": 1.306,
+      "grad_norm": 79.32563781738281,
+      "learning_rate": 4.778319999999999e-07,
+      "loss": 0.3935,
+      "step": 130600
+    },
+    {
+      "epoch": 1.3065,
+      "grad_norm": 12.994901657104492,
+      "learning_rate": 4.77632e-07,
+      "loss": 0.4861,
+      "step": 130650
+    },
+    {
+      "epoch": 1.307,
+      "grad_norm": 34.566383361816406,
+      "learning_rate": 4.77432e-07,
+      "loss": 0.4927,
+      "step": 130700
+    },
+    {
+      "epoch": 1.3075,
+      "grad_norm": 80.90245056152344,
+      "learning_rate": 4.77232e-07,
+      "loss": 0.5533,
+      "step": 130750
+    },
+    {
+      "epoch": 1.308,
+      "grad_norm": 7.320152282714844,
+      "learning_rate": 4.77032e-07,
+      "loss": 0.4052,
+      "step": 130800
+    },
+    {
+      "epoch": 1.3085,
+      "grad_norm": 9.89566421508789,
+      "learning_rate": 4.7683199999999996e-07,
+      "loss": 0.411,
+      "step": 130850
+    },
+    {
+      "epoch": 1.309,
+      "grad_norm": 32.64478302001953,
+      "learning_rate": 4.76632e-07,
+      "loss": 0.4795,
+      "step": 130900
+    },
+    {
+      "epoch": 1.3094999999999999,
+      "grad_norm": 4.066441059112549,
+      "learning_rate": 4.76432e-07,
+      "loss": 0.637,
+      "step": 130950
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 147.58811950683594,
+      "learning_rate": 4.7623199999999997e-07,
+      "loss": 0.6026,
+      "step": 131000
+    },
+    {
+      "epoch": 1.3105,
+      "grad_norm": 11.256807327270508,
+      "learning_rate": 4.76032e-07,
+      "loss": 0.5771,
+      "step": 131050
+    },
+    {
+      "epoch": 1.311,
+      "grad_norm": 52.247196197509766,
+      "learning_rate": 4.7583199999999994e-07,
+      "loss": 0.4587,
+      "step": 131100
+    },
+    {
+      "epoch": 1.3115,
+      "grad_norm": 1.7685823440551758,
+      "learning_rate": 4.75632e-07,
+      "loss": 0.3916,
+      "step": 131150
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 72.10371398925781,
+      "learning_rate": 4.75432e-07,
+      "loss": 0.3049,
+      "step": 131200
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 86.47714233398438,
+      "learning_rate": 4.7523199999999995e-07,
+      "loss": 0.4716,
+      "step": 131250
+    },
+    {
+      "epoch": 1.313,
+      "grad_norm": 120.87084197998047,
+      "learning_rate": 4.75032e-07,
+      "loss": 0.4428,
+      "step": 131300
+    },
+    {
+      "epoch": 1.3135,
+      "grad_norm": 21.09920883178711,
+      "learning_rate": 4.7483199999999997e-07,
+      "loss": 0.521,
+      "step": 131350
+    },
+    {
+      "epoch": 1.314,
+      "grad_norm": 88.9284439086914,
+      "learning_rate": 4.7463199999999996e-07,
+      "loss": 0.5036,
+      "step": 131400
+    },
+    {
+      "epoch": 1.3145,
+      "grad_norm": 56.702247619628906,
+      "learning_rate": 4.74432e-07,
+      "loss": 0.5041,
+      "step": 131450
+    },
+    {
+      "epoch": 1.315,
+      "grad_norm": 0.35143765807151794,
+      "learning_rate": 4.74232e-07,
+      "loss": 0.4163,
+      "step": 131500
+    },
+    {
+      "epoch": 1.3155000000000001,
+      "grad_norm": 92.69734191894531,
+      "learning_rate": 4.7403199999999997e-07,
+      "loss": 0.4294,
+      "step": 131550
+    },
+    {
+      "epoch": 1.316,
+      "grad_norm": 50.940093994140625,
+      "learning_rate": 4.7383199999999995e-07,
+      "loss": 0.4826,
+      "step": 131600
+    },
+    {
+      "epoch": 1.3165,
+      "grad_norm": 82.14228820800781,
+      "learning_rate": 4.73632e-07,
+      "loss": 0.4999,
+      "step": 131650
+    },
+    {
+      "epoch": 1.317,
+      "grad_norm": 75.90847778320312,
+      "learning_rate": 4.73432e-07,
+      "loss": 0.4275,
+      "step": 131700
+    },
+    {
+      "epoch": 1.3175,
+      "grad_norm": 34.303688049316406,
+      "learning_rate": 4.7323199999999996e-07,
+      "loss": 0.453,
+      "step": 131750
+    },
+    {
+      "epoch": 1.318,
+      "grad_norm": 47.25891876220703,
+      "learning_rate": 4.73032e-07,
+      "loss": 0.5394,
+      "step": 131800
+    },
+    {
+      "epoch": 1.3185,
+      "grad_norm": 0.128638356924057,
+      "learning_rate": 4.7283199999999993e-07,
+      "loss": 0.3947,
+      "step": 131850
+    },
+    {
+      "epoch": 1.319,
+      "grad_norm": 4.6934967041015625,
+      "learning_rate": 4.7263199999999997e-07,
+      "loss": 0.5127,
+      "step": 131900
+    },
+    {
+      "epoch": 1.3195000000000001,
+      "grad_norm": 38.059913635253906,
+      "learning_rate": 4.72432e-07,
+      "loss": 0.4648,
+      "step": 131950
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 42.113609313964844,
+      "learning_rate": 4.72232e-07,
+      "loss": 0.3855,
+      "step": 132000
+    },
+    {
+      "epoch": 1.3205,
+      "grad_norm": 88.45838928222656,
+      "learning_rate": 4.72032e-07,
+      "loss": 0.3796,
+      "step": 132050
+    },
+    {
+      "epoch": 1.321,
+      "grad_norm": 60.03329086303711,
+      "learning_rate": 4.7183199999999996e-07,
+      "loss": 0.4678,
+      "step": 132100
+    },
+    {
+      "epoch": 1.3215,
+      "grad_norm": 92.38176727294922,
+      "learning_rate": 4.71632e-07,
+      "loss": 0.4438,
+      "step": 132150
+    },
+    {
+      "epoch": 1.322,
+      "grad_norm": 5.0574235916137695,
+      "learning_rate": 4.71432e-07,
+      "loss": 0.4548,
+      "step": 132200
+    },
+    {
+      "epoch": 1.3225,
+      "grad_norm": 45.51777648925781,
+      "learning_rate": 4.7123199999999997e-07,
+      "loss": 0.477,
+      "step": 132250
+    },
+    {
+      "epoch": 1.323,
+      "grad_norm": 58.61941146850586,
+      "learning_rate": 4.71032e-07,
+      "loss": 0.4616,
+      "step": 132300
+    },
+    {
+      "epoch": 1.3235000000000001,
+      "grad_norm": 9.93522834777832,
+      "learning_rate": 4.7083199999999994e-07,
+      "loss": 0.3566,
+      "step": 132350
+    },
+    {
+      "epoch": 1.324,
+      "grad_norm": 73.70626068115234,
+      "learning_rate": 4.70632e-07,
+      "loss": 0.5039,
+      "step": 132400
+    },
+    {
+      "epoch": 1.3245,
+      "grad_norm": 119.5567626953125,
+      "learning_rate": 4.70432e-07,
+      "loss": 0.5233,
+      "step": 132450
+    },
+    {
+      "epoch": 1.325,
+      "grad_norm": 93.0530014038086,
+      "learning_rate": 4.7023199999999995e-07,
+      "loss": 0.3723,
+      "step": 132500
+    },
+    {
+      "epoch": 1.3255,
+      "grad_norm": 94.75311279296875,
+      "learning_rate": 4.70032e-07,
+      "loss": 0.4238,
+      "step": 132550
+    },
+    {
+      "epoch": 1.326,
+      "grad_norm": 24.6379337310791,
+      "learning_rate": 4.69832e-07,
+      "loss": 0.5475,
+      "step": 132600
+    },
+    {
+      "epoch": 1.3265,
+      "grad_norm": 128.80043029785156,
+      "learning_rate": 4.6963199999999995e-07,
+      "loss": 0.5309,
+      "step": 132650
+    },
+    {
+      "epoch": 1.327,
+      "grad_norm": 11.744426727294922,
+      "learning_rate": 4.69432e-07,
+      "loss": 0.4684,
+      "step": 132700
+    },
+    {
+      "epoch": 1.3275000000000001,
+      "grad_norm": 68.0375747680664,
+      "learning_rate": 4.69232e-07,
+      "loss": 0.4713,
+      "step": 132750
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 63.520179748535156,
+      "learning_rate": 4.6903199999999996e-07,
+      "loss": 0.4248,
+      "step": 132800
+    },
+    {
+      "epoch": 1.3285,
+      "grad_norm": 10.20051383972168,
+      "learning_rate": 4.68832e-07,
+      "loss": 0.3877,
+      "step": 132850
+    },
+    {
+      "epoch": 1.329,
+      "grad_norm": 94.5232925415039,
+      "learning_rate": 4.68632e-07,
+      "loss": 0.5337,
+      "step": 132900
+    },
+    {
+      "epoch": 1.3295,
+      "grad_norm": 118.54349517822266,
+      "learning_rate": 4.6843199999999997e-07,
+      "loss": 0.3081,
+      "step": 132950
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.4136791229248047,
+      "learning_rate": 4.6823199999999995e-07,
+      "loss": 0.4322,
+      "step": 133000
+    },
+    {
+      "epoch": 1.3305,
+      "grad_norm": 13.606802940368652,
+      "learning_rate": 4.68032e-07,
+      "loss": 0.5477,
+      "step": 133050
+    },
+    {
+      "epoch": 1.331,
+      "grad_norm": 58.811222076416016,
+      "learning_rate": 4.6783200000000003e-07,
+      "loss": 0.5183,
+      "step": 133100
+    },
+    {
+      "epoch": 1.3315000000000001,
+      "grad_norm": 1.774475336074829,
+      "learning_rate": 4.6763199999999996e-07,
+      "loss": 0.3814,
+      "step": 133150
+    },
+    {
+      "epoch": 1.332,
+      "grad_norm": 3.9559104442596436,
+      "learning_rate": 4.67432e-07,
+      "loss": 0.4848,
+      "step": 133200
+    },
+    {
+      "epoch": 1.3325,
+      "grad_norm": 44.5838623046875,
+      "learning_rate": 4.67232e-07,
+      "loss": 0.4987,
+      "step": 133250
+    },
+    {
+      "epoch": 1.333,
+      "grad_norm": 45.46150207519531,
+      "learning_rate": 4.6703199999999997e-07,
+      "loss": 0.4701,
+      "step": 133300
+    },
+    {
+      "epoch": 1.3335,
+      "grad_norm": 4.442442417144775,
+      "learning_rate": 4.66832e-07,
+      "loss": 0.4129,
+      "step": 133350
+    },
+    {
+      "epoch": 1.334,
+      "grad_norm": 83.65077209472656,
+      "learning_rate": 4.66632e-07,
+      "loss": 0.4445,
+      "step": 133400
+    },
+    {
+      "epoch": 1.3345,
+      "grad_norm": 95.16340637207031,
+      "learning_rate": 4.66432e-07,
+      "loss": 0.4481,
+      "step": 133450
+    },
+    {
+      "epoch": 1.335,
+      "grad_norm": 12.23143482208252,
+      "learning_rate": 4.6623199999999996e-07,
+      "loss": 0.4512,
+      "step": 133500
+    },
+    {
+      "epoch": 1.3355000000000001,
+      "grad_norm": 38.25416564941406,
+      "learning_rate": 4.66032e-07,
+      "loss": 0.3672,
+      "step": 133550
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 57.15376281738281,
+      "learning_rate": 4.65832e-07,
+      "loss": 0.5406,
+      "step": 133600
+    },
+    {
+      "epoch": 1.3365,
+      "grad_norm": 58.421695709228516,
+      "learning_rate": 4.6563199999999997e-07,
+      "loss": 0.5141,
+      "step": 133650
+    },
+    {
+      "epoch": 1.337,
+      "grad_norm": 87.27682495117188,
+      "learning_rate": 4.65432e-07,
+      "loss": 0.4306,
+      "step": 133700
+    },
+    {
+      "epoch": 1.3375,
+      "grad_norm": 71.72251892089844,
+      "learning_rate": 4.6523199999999994e-07,
+      "loss": 0.4374,
+      "step": 133750
+    },
+    {
+      "epoch": 1.338,
+      "grad_norm": 49.080322265625,
+      "learning_rate": 4.65032e-07,
+      "loss": 0.5072,
+      "step": 133800
+    },
+    {
+      "epoch": 1.3385,
+      "grad_norm": 22.312482833862305,
+      "learning_rate": 4.64832e-07,
+      "loss": 0.4124,
+      "step": 133850
+    },
+    {
+      "epoch": 1.339,
+      "grad_norm": 36.95961380004883,
+      "learning_rate": 4.6463199999999995e-07,
+      "loss": 0.4986,
+      "step": 133900
+    },
+    {
+      "epoch": 1.3395000000000001,
+      "grad_norm": 2.186569929122925,
+      "learning_rate": 4.64432e-07,
+      "loss": 0.3585,
+      "step": 133950
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 38.58488464355469,
+      "learning_rate": 4.6423199999999997e-07,
+      "loss": 0.4046,
+      "step": 134000
+    },
+    {
+      "epoch": 1.3405,
+      "grad_norm": 83.42021942138672,
+      "learning_rate": 4.6403199999999996e-07,
+      "loss": 0.4514,
+      "step": 134050
+    },
+    {
+      "epoch": 1.341,
+      "grad_norm": 5.808516979217529,
+      "learning_rate": 4.63832e-07,
+      "loss": 0.4749,
+      "step": 134100
+    },
+    {
+      "epoch": 1.3415,
+      "grad_norm": 6.701382637023926,
+      "learning_rate": 4.63632e-07,
+      "loss": 0.4058,
+      "step": 134150
+    },
+    {
+      "epoch": 1.342,
+      "grad_norm": 54.70067596435547,
+      "learning_rate": 4.6343199999999996e-07,
+      "loss": 0.4774,
+      "step": 134200
+    },
+    {
+      "epoch": 1.3425,
+      "grad_norm": 0.1387663632631302,
+      "learning_rate": 4.6323199999999995e-07,
+      "loss": 0.3653,
+      "step": 134250
+    },
+    {
+      "epoch": 1.343,
+      "grad_norm": 32.30421829223633,
+      "learning_rate": 4.63032e-07,
+      "loss": 0.3759,
+      "step": 134300
+    },
+    {
+      "epoch": 1.3435000000000001,
+      "grad_norm": 60.436344146728516,
+      "learning_rate": 4.62832e-07,
+      "loss": 0.5095,
+      "step": 134350
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 83.42176818847656,
+      "learning_rate": 4.6263199999999996e-07,
+      "loss": 0.425,
+      "step": 134400
+    },
+    {
+      "epoch": 1.3445,
+      "grad_norm": 3.811023473739624,
+      "learning_rate": 4.62432e-07,
+      "loss": 0.407,
+      "step": 134450
+    },
+    {
+      "epoch": 1.345,
+      "grad_norm": 74.2744369506836,
+      "learning_rate": 4.62232e-07,
+      "loss": 0.4088,
+      "step": 134500
+    },
+    {
+      "epoch": 1.3455,
+      "grad_norm": 7.530820846557617,
+      "learning_rate": 4.6203199999999996e-07,
+      "loss": 0.401,
+      "step": 134550
+    },
+    {
+      "epoch": 1.346,
+      "grad_norm": 143.85891723632812,
+      "learning_rate": 4.61832e-07,
+      "loss": 0.5301,
+      "step": 134600
+    },
+    {
+      "epoch": 1.3465,
+      "grad_norm": 82.85350036621094,
+      "learning_rate": 4.61632e-07,
+      "loss": 0.4794,
+      "step": 134650
+    },
+    {
+      "epoch": 1.347,
+      "grad_norm": 94.02362823486328,
+      "learning_rate": 4.6143199999999997e-07,
+      "loss": 0.408,
+      "step": 134700
+    },
+    {
+      "epoch": 1.3475,
+      "grad_norm": 1.712831974029541,
+      "learning_rate": 4.6123199999999996e-07,
+      "loss": 0.3879,
+      "step": 134750
+    },
+    {
+      "epoch": 1.3479999999999999,
+      "grad_norm": 91.60545349121094,
+      "learning_rate": 4.61032e-07,
+      "loss": 0.4832,
+      "step": 134800
+    },
+    {
+      "epoch": 1.3485,
+      "grad_norm": 107.8651123046875,
+      "learning_rate": 4.60832e-07,
+      "loss": 0.4833,
+      "step": 134850
+    },
+    {
+      "epoch": 1.349,
+      "grad_norm": 70.89888000488281,
+      "learning_rate": 4.6063199999999997e-07,
+      "loss": 0.4173,
+      "step": 134900
+    },
+    {
+      "epoch": 1.3495,
+      "grad_norm": 100.18441772460938,
+      "learning_rate": 4.60432e-07,
+      "loss": 0.5035,
+      "step": 134950
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.220166206359863,
+      "learning_rate": 4.6023599999999994e-07,
+      "loss": 0.3394,
+      "step": 135000
+    },
+    {
+      "epoch": 1.3505,
+      "grad_norm": 76.76039123535156,
+      "learning_rate": 4.60036e-07,
+      "loss": 0.4507,
+      "step": 135050
+    },
+    {
+      "epoch": 1.351,
+      "grad_norm": 94.28003692626953,
+      "learning_rate": 4.59836e-07,
+      "loss": 0.3727,
+      "step": 135100
+    },
+    {
+      "epoch": 1.3515,
+      "grad_norm": 109.33968353271484,
+      "learning_rate": 4.5963599999999995e-07,
+      "loss": 0.4583,
+      "step": 135150
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 50.13473892211914,
+      "learning_rate": 4.59436e-07,
+      "loss": 0.4678,
+      "step": 135200
+    },
+    {
+      "epoch": 1.3525,
+      "grad_norm": 30.904821395874023,
+      "learning_rate": 4.5923599999999997e-07,
+      "loss": 0.4636,
+      "step": 135250
+    },
+    {
+      "epoch": 1.353,
+      "grad_norm": 115.35250091552734,
+      "learning_rate": 4.5903599999999996e-07,
+      "loss": 0.3913,
+      "step": 135300
+    },
+    {
+      "epoch": 1.3535,
+      "grad_norm": 116.52228546142578,
+      "learning_rate": 4.58836e-07,
+      "loss": 0.553,
+      "step": 135350
+    },
+    {
+      "epoch": 1.354,
+      "grad_norm": 86.03593444824219,
+      "learning_rate": 4.58636e-07,
+      "loss": 0.5031,
+      "step": 135400
+    },
+    {
+      "epoch": 1.3545,
+      "grad_norm": 72.3880386352539,
+      "learning_rate": 4.5843599999999996e-07,
+      "loss": 0.4295,
+      "step": 135450
+    },
+    {
+      "epoch": 1.355,
+      "grad_norm": 45.91992950439453,
+      "learning_rate": 4.5823599999999995e-07,
+      "loss": 0.396,
+      "step": 135500
+    },
+    {
+      "epoch": 1.3555,
+      "grad_norm": 17.957754135131836,
+      "learning_rate": 4.58036e-07,
+      "loss": 0.3191,
+      "step": 135550
+    },
+    {
+      "epoch": 1.3559999999999999,
+      "grad_norm": 3.7718541622161865,
+      "learning_rate": 4.57836e-07,
+      "loss": 0.5687,
+      "step": 135600
+    },
+    {
+      "epoch": 1.3565,
+      "grad_norm": 16.24213409423828,
+      "learning_rate": 4.5763599999999996e-07,
+      "loss": 0.4716,
+      "step": 135650
+    },
+    {
+      "epoch": 1.357,
+      "grad_norm": 3.7620809078216553,
+      "learning_rate": 4.57436e-07,
+      "loss": 0.456,
+      "step": 135700
+    },
+    {
+      "epoch": 1.3575,
+      "grad_norm": 86.84635925292969,
+      "learning_rate": 4.57236e-07,
+      "loss": 0.3689,
+      "step": 135750
+    },
+    {
+      "epoch": 1.358,
+      "grad_norm": 84.04347229003906,
+      "learning_rate": 4.5703599999999996e-07,
+      "loss": 0.4508,
+      "step": 135800
+    },
+    {
+      "epoch": 1.3585,
+      "grad_norm": 29.693422317504883,
+      "learning_rate": 4.56836e-07,
+      "loss": 0.5614,
+      "step": 135850
+    },
+    {
+      "epoch": 1.359,
+      "grad_norm": 45.12217712402344,
+      "learning_rate": 4.56636e-07,
+      "loss": 0.3916,
+      "step": 135900
+    },
+    {
+      "epoch": 1.3595,
+      "grad_norm": 31.29984474182129,
+      "learning_rate": 4.5643599999999997e-07,
+      "loss": 0.4284,
+      "step": 135950
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 33.480751037597656,
+      "learning_rate": 4.5623599999999996e-07,
+      "loss": 0.3547,
+      "step": 136000
+    },
+    {
+      "epoch": 1.3605,
+      "grad_norm": 2.545854091644287,
+      "learning_rate": 4.56036e-07,
+      "loss": 0.418,
+      "step": 136050
+    },
+    {
+      "epoch": 1.361,
+      "grad_norm": 91.34917449951172,
+      "learning_rate": 4.55836e-07,
+      "loss": 0.4277,
+      "step": 136100
+    },
+    {
+      "epoch": 1.3615,
+      "grad_norm": 36.96833419799805,
+      "learning_rate": 4.5563599999999996e-07,
+      "loss": 0.4769,
+      "step": 136150
+    },
+    {
+      "epoch": 1.362,
+      "grad_norm": 53.53211975097656,
+      "learning_rate": 4.55436e-07,
+      "loss": 0.4037,
+      "step": 136200
+    },
+    {
+      "epoch": 1.3625,
+      "grad_norm": 16.555437088012695,
+      "learning_rate": 4.5523599999999993e-07,
+      "loss": 0.4622,
+      "step": 136250
+    },
+    {
+      "epoch": 1.363,
+      "grad_norm": 2.20904803276062,
+      "learning_rate": 4.5503599999999997e-07,
+      "loss": 0.4856,
+      "step": 136300
+    },
+    {
+      "epoch": 1.3635,
+      "grad_norm": 67.71829986572266,
+      "learning_rate": 4.54836e-07,
+      "loss": 0.4294,
+      "step": 136350
+    },
+    {
+      "epoch": 1.3639999999999999,
+      "grad_norm": 27.427248001098633,
+      "learning_rate": 4.5463599999999994e-07,
+      "loss": 0.5167,
+      "step": 136400
+    },
+    {
+      "epoch": 1.3645,
+      "grad_norm": 80.44205474853516,
+      "learning_rate": 4.54436e-07,
+      "loss": 0.4855,
+      "step": 136450
+    },
+    {
+      "epoch": 1.365,
+      "grad_norm": 15.163834571838379,
+      "learning_rate": 4.5423599999999997e-07,
+      "loss": 0.3511,
+      "step": 136500
+    },
+    {
+      "epoch": 1.3655,
+      "grad_norm": 12.879931449890137,
+      "learning_rate": 4.5403599999999995e-07,
+      "loss": 0.4367,
+      "step": 136550
+    },
+    {
+      "epoch": 1.366,
+      "grad_norm": 68.53999328613281,
+      "learning_rate": 4.53836e-07,
+      "loss": 0.5996,
+      "step": 136600
+    },
+    {
+      "epoch": 1.3665,
+      "grad_norm": 121.40241241455078,
+      "learning_rate": 4.5363599999999997e-07,
+      "loss": 0.5384,
+      "step": 136650
+    },
+    {
+      "epoch": 1.367,
+      "grad_norm": 0.31804296374320984,
+      "learning_rate": 4.53436e-07,
+      "loss": 0.6281,
+      "step": 136700
+    },
+    {
+      "epoch": 1.3675,
+      "grad_norm": 13.64835262298584,
+      "learning_rate": 4.53236e-07,
+      "loss": 0.4431,
+      "step": 136750
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 39.9966926574707,
+      "learning_rate": 4.53036e-07,
+      "loss": 0.4694,
+      "step": 136800
+    },
+    {
+      "epoch": 1.3685,
+      "grad_norm": 29.257156372070312,
+      "learning_rate": 4.52836e-07,
+      "loss": 0.5316,
+      "step": 136850
+    },
+    {
+      "epoch": 1.369,
+      "grad_norm": 68.866943359375,
+      "learning_rate": 4.5263599999999995e-07,
+      "loss": 0.4085,
+      "step": 136900
+    },
+    {
+      "epoch": 1.3695,
+      "grad_norm": 1.851599097251892,
+      "learning_rate": 4.52436e-07,
+      "loss": 0.4832,
+      "step": 136950
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 66.2961654663086,
+      "learning_rate": 4.5223600000000003e-07,
+      "loss": 0.4054,
+      "step": 137000
+    },
+    {
+      "epoch": 1.3705,
+      "grad_norm": 12.196954727172852,
+      "learning_rate": 4.5203599999999996e-07,
+      "loss": 0.5535,
+      "step": 137050
+    },
+    {
+      "epoch": 1.371,
+      "grad_norm": 60.95892333984375,
+      "learning_rate": 4.51836e-07,
+      "loss": 0.3948,
+      "step": 137100
+    },
+    {
+      "epoch": 1.3715,
+      "grad_norm": 0.48031020164489746,
+      "learning_rate": 4.51636e-07,
+      "loss": 0.5336,
+      "step": 137150
+    },
+    {
+      "epoch": 1.3719999999999999,
+      "grad_norm": 88.17898559570312,
+      "learning_rate": 4.5143599999999997e-07,
+      "loss": 0.553,
+      "step": 137200
+    },
+    {
+      "epoch": 1.3725,
+      "grad_norm": 15.11093807220459,
+      "learning_rate": 4.51236e-07,
+      "loss": 0.4888,
+      "step": 137250
+    },
+    {
+      "epoch": 1.373,
+      "grad_norm": 27.76417350769043,
+      "learning_rate": 4.51036e-07,
+      "loss": 0.5925,
+      "step": 137300
+    },
+    {
+      "epoch": 1.3735,
+      "grad_norm": 2.448547840118408,
+      "learning_rate": 4.50836e-07,
+      "loss": 0.3783,
+      "step": 137350
+    },
+    {
+      "epoch": 1.374,
+      "grad_norm": 7.2902607917785645,
+      "learning_rate": 4.5063599999999996e-07,
+      "loss": 0.4221,
+      "step": 137400
+    },
+    {
+      "epoch": 1.3745,
+      "grad_norm": 15.440958976745605,
+      "learning_rate": 4.50436e-07,
+      "loss": 0.3223,
+      "step": 137450
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 9.393592834472656,
+      "learning_rate": 4.50236e-07,
+      "loss": 0.393,
+      "step": 137500
+    },
+    {
+      "epoch": 1.3755,
+      "grad_norm": 22.066003799438477,
+      "learning_rate": 4.5003599999999997e-07,
+      "loss": 0.4808,
+      "step": 137550
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 2.079468250274658,
+      "learning_rate": 4.49836e-07,
+      "loss": 0.4312,
+      "step": 137600
+    },
+    {
+      "epoch": 1.3765,
+      "grad_norm": 74.9188461303711,
+      "learning_rate": 4.4963599999999994e-07,
+      "loss": 0.4377,
+      "step": 137650
+    },
+    {
+      "epoch": 1.377,
+      "grad_norm": 0.9050947427749634,
+      "learning_rate": 4.49436e-07,
+      "loss": 0.4035,
+      "step": 137700
+    },
+    {
+      "epoch": 1.3775,
+      "grad_norm": 104.04845428466797,
+      "learning_rate": 4.49236e-07,
+      "loss": 0.3991,
+      "step": 137750
+    },
+    {
+      "epoch": 1.3780000000000001,
+      "grad_norm": 47.782188415527344,
+      "learning_rate": 4.4903599999999995e-07,
+      "loss": 0.4328,
+      "step": 137800
+    },
+    {
+      "epoch": 1.3785,
+      "grad_norm": 66.12909698486328,
+      "learning_rate": 4.48836e-07,
+      "loss": 0.4564,
+      "step": 137850
+    },
+    {
+      "epoch": 1.379,
+      "grad_norm": 99.37031555175781,
+      "learning_rate": 4.4863599999999997e-07,
+      "loss": 0.4904,
+      "step": 137900
+    },
+    {
+      "epoch": 1.3795,
+      "grad_norm": 110.17735290527344,
+      "learning_rate": 4.48436e-07,
+      "loss": 0.5182,
+      "step": 137950
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 50.46833038330078,
+      "learning_rate": 4.48236e-07,
+      "loss": 0.5215,
+      "step": 138000
+    },
+    {
+      "epoch": 1.3805,
+      "grad_norm": 110.3285140991211,
+      "learning_rate": 4.48036e-07,
+      "loss": 0.5516,
+      "step": 138050
+    },
+    {
+      "epoch": 1.381,
+      "grad_norm": 1.3695465326309204,
+      "learning_rate": 4.47836e-07,
+      "loss": 0.4337,
+      "step": 138100
+    },
+    {
+      "epoch": 1.3815,
+      "grad_norm": 38.30186462402344,
+      "learning_rate": 4.4763599999999995e-07,
+      "loss": 0.4659,
+      "step": 138150
+    },
+    {
+      "epoch": 1.3820000000000001,
+      "grad_norm": 70.28520202636719,
+      "learning_rate": 4.47436e-07,
+      "loss": 0.48,
+      "step": 138200
+    },
+    {
+      "epoch": 1.3825,
+      "grad_norm": 9.3742036819458,
+      "learning_rate": 4.47236e-07,
+      "loss": 0.4788,
+      "step": 138250
+    },
+    {
+      "epoch": 1.383,
+      "grad_norm": 0.6705490350723267,
+      "learning_rate": 4.4703599999999995e-07,
+      "loss": 0.4825,
+      "step": 138300
+    },
+    {
+      "epoch": 1.3835,
+      "grad_norm": 30.913881301879883,
+      "learning_rate": 4.46836e-07,
+      "loss": 0.4678,
+      "step": 138350
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 71.79946899414062,
+      "learning_rate": 4.46636e-07,
+      "loss": 0.5039,
+      "step": 138400
+    },
+    {
+      "epoch": 1.3845,
+      "grad_norm": 25.541101455688477,
+      "learning_rate": 4.4643599999999996e-07,
+      "loss": 0.4999,
+      "step": 138450
+    },
+    {
+      "epoch": 1.385,
+      "grad_norm": 88.34212493896484,
+      "learning_rate": 4.46236e-07,
+      "loss": 0.3772,
+      "step": 138500
+    },
+    {
+      "epoch": 1.3855,
+      "grad_norm": 28.59035873413086,
+      "learning_rate": 4.46036e-07,
+      "loss": 0.3627,
+      "step": 138550
+    },
+    {
+      "epoch": 1.3860000000000001,
+      "grad_norm": 60.05609130859375,
+      "learning_rate": 4.4583599999999997e-07,
+      "loss": 0.458,
+      "step": 138600
+    },
+    {
+      "epoch": 1.3865,
+      "grad_norm": 52.41706085205078,
+      "learning_rate": 4.4563599999999995e-07,
+      "loss": 0.3964,
+      "step": 138650
+    },
+    {
+      "epoch": 1.387,
+      "grad_norm": 121.22614288330078,
+      "learning_rate": 4.45436e-07,
+      "loss": 0.3593,
+      "step": 138700
+    },
+    {
+      "epoch": 1.3875,
+      "grad_norm": 9.27535629272461,
+      "learning_rate": 4.45236e-07,
+      "loss": 0.5142,
+      "step": 138750
+    },
+    {
+      "epoch": 1.388,
+      "grad_norm": 5.494357109069824,
+      "learning_rate": 4.4503599999999996e-07,
+      "loss": 0.4425,
+      "step": 138800
+    },
+    {
+      "epoch": 1.3885,
+      "grad_norm": 100.4722671508789,
+      "learning_rate": 4.44836e-07,
+      "loss": 0.4553,
+      "step": 138850
+    },
+    {
+      "epoch": 1.389,
+      "grad_norm": 6.942700386047363,
+      "learning_rate": 4.4463599999999993e-07,
+      "loss": 0.4525,
+      "step": 138900
+    },
+    {
+      "epoch": 1.3895,
+      "grad_norm": 38.53964614868164,
+      "learning_rate": 4.4443599999999997e-07,
+      "loss": 0.5523,
+      "step": 138950
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 38.89802169799805,
+      "learning_rate": 4.44236e-07,
+      "loss": 0.3422,
+      "step": 139000
+    },
+    {
+      "epoch": 1.3905,
+      "grad_norm": 123.1884765625,
+      "learning_rate": 4.4403599999999994e-07,
+      "loss": 0.6059,
+      "step": 139050
+    },
+    {
+      "epoch": 1.391,
+      "grad_norm": 3.0012309551239014,
+      "learning_rate": 4.43836e-07,
+      "loss": 0.3437,
+      "step": 139100
+    },
+    {
+      "epoch": 1.3915,
+      "grad_norm": 1.8544474840164185,
+      "learning_rate": 4.4363599999999996e-07,
+      "loss": 0.3483,
+      "step": 139150
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 81.5112533569336,
+      "learning_rate": 4.43436e-07,
+      "loss": 0.4729,
+      "step": 139200
+    },
+    {
+      "epoch": 1.3925,
+      "grad_norm": 88.00301361083984,
+      "learning_rate": 4.43236e-07,
+      "loss": 0.4897,
+      "step": 139250
+    },
+    {
+      "epoch": 1.393,
+      "grad_norm": 7.775111675262451,
+      "learning_rate": 4.4303599999999997e-07,
+      "loss": 0.4474,
+      "step": 139300
+    },
+    {
+      "epoch": 1.3935,
+      "grad_norm": 60.3265266418457,
+      "learning_rate": 4.42836e-07,
+      "loss": 0.5121,
+      "step": 139350
+    },
+    {
+      "epoch": 1.3940000000000001,
+      "grad_norm": 79.7757568359375,
+      "learning_rate": 4.4263599999999994e-07,
+      "loss": 0.512,
+      "step": 139400
+    },
+    {
+      "epoch": 1.3945,
+      "grad_norm": 13.68470287322998,
+      "learning_rate": 4.42436e-07,
+      "loss": 0.3414,
+      "step": 139450
+    },
+    {
+      "epoch": 1.395,
+      "grad_norm": 60.95335388183594,
+      "learning_rate": 4.42236e-07,
+      "loss": 0.4045,
+      "step": 139500
+    },
+    {
+      "epoch": 1.3955,
+      "grad_norm": 47.811424255371094,
+      "learning_rate": 4.4203599999999995e-07,
+      "loss": 0.384,
+      "step": 139550
+    },
+    {
+      "epoch": 1.396,
+      "grad_norm": 17.77873420715332,
+      "learning_rate": 4.41836e-07,
+      "loss": 0.4314,
+      "step": 139600
+    },
+    {
+      "epoch": 1.3965,
+      "grad_norm": 0.8219741582870483,
+      "learning_rate": 4.4163599999999997e-07,
+      "loss": 0.4474,
+      "step": 139650
+    },
+    {
+      "epoch": 1.397,
+      "grad_norm": 44.76259231567383,
+      "learning_rate": 4.4143599999999996e-07,
+      "loss": 0.4728,
+      "step": 139700
+    },
+    {
+      "epoch": 1.3975,
+      "grad_norm": 32.0296745300293,
+      "learning_rate": 4.41236e-07,
+      "loss": 0.2824,
+      "step": 139750
+    },
+    {
+      "epoch": 1.3980000000000001,
+      "grad_norm": 6.995255947113037,
+      "learning_rate": 4.41036e-07,
+      "loss": 0.4395,
+      "step": 139800
+    },
+    {
+      "epoch": 1.3985,
+      "grad_norm": 3.6521756649017334,
+      "learning_rate": 4.4083599999999996e-07,
+      "loss": 0.4282,
+      "step": 139850
+    },
+    {
+      "epoch": 1.399,
+      "grad_norm": 0.5228773951530457,
+      "learning_rate": 4.4063999999999995e-07,
+      "loss": 0.5437,
+      "step": 139900
+    },
+    {
+      "epoch": 1.3995,
+      "grad_norm": 0.48555630445480347,
+      "learning_rate": 4.4044e-07,
+      "loss": 0.5651,
+      "step": 139950
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 81.31806182861328,
+      "learning_rate": 4.4024e-07,
+      "loss": 0.4659,
+      "step": 140000
+    },
+    {
+      "epoch": 1.4005,
+      "grad_norm": 1.3937146663665771,
+      "learning_rate": 4.4003999999999996e-07,
+      "loss": 0.4486,
+      "step": 140050
+    },
+    {
+      "epoch": 1.401,
+      "grad_norm": 21.106035232543945,
+      "learning_rate": 4.3984e-07,
+      "loss": 0.4779,
+      "step": 140100
+    },
+    {
+      "epoch": 1.4015,
+      "grad_norm": 13.850019454956055,
+      "learning_rate": 4.3963999999999993e-07,
+      "loss": 0.3871,
+      "step": 140150
+    },
+    {
+      "epoch": 1.4020000000000001,
+      "grad_norm": 1.2949334383010864,
+      "learning_rate": 4.3943999999999997e-07,
+      "loss": 0.5905,
+      "step": 140200
+    },
+    {
+      "epoch": 1.4025,
+      "grad_norm": 85.25445556640625,
+      "learning_rate": 4.3924e-07,
+      "loss": 0.5164,
+      "step": 140250
+    },
+    {
+      "epoch": 1.403,
+      "grad_norm": 21.79319190979004,
+      "learning_rate": 4.3903999999999994e-07,
+      "loss": 0.4014,
+      "step": 140300
+    },
+    {
+      "epoch": 1.4035,
+      "grad_norm": 16.37718391418457,
+      "learning_rate": 4.3884e-07,
+      "loss": 0.4419,
+      "step": 140350
+    },
+    {
+      "epoch": 1.404,
+      "grad_norm": 3.0240814685821533,
+      "learning_rate": 4.3863999999999996e-07,
+      "loss": 0.5172,
+      "step": 140400
+    },
+    {
+      "epoch": 1.4045,
+      "grad_norm": 106.63111877441406,
+      "learning_rate": 4.3844e-07,
+      "loss": 0.4104,
+      "step": 140450
+    },
+    {
+      "epoch": 1.405,
+      "grad_norm": 17.3591251373291,
+      "learning_rate": 4.3824e-07,
+      "loss": 0.4768,
+      "step": 140500
+    },
+    {
+      "epoch": 1.4055,
+      "grad_norm": 76.1080322265625,
+      "learning_rate": 4.3803999999999997e-07,
+      "loss": 0.4747,
+      "step": 140550
+    },
+    {
+      "epoch": 1.4060000000000001,
+      "grad_norm": 3.34743595123291,
+      "learning_rate": 4.3784e-07,
+      "loss": 0.3191,
+      "step": 140600
+    },
+    {
+      "epoch": 1.4064999999999999,
+      "grad_norm": 38.74766159057617,
+      "learning_rate": 4.3763999999999994e-07,
+      "loss": 0.4569,
+      "step": 140650
+    },
+    {
+      "epoch": 1.407,
+      "grad_norm": 46.4010124206543,
+      "learning_rate": 4.3744e-07,
+      "loss": 0.3989,
+      "step": 140700
+    },
+    {
+      "epoch": 1.4075,
+      "grad_norm": 69.19367980957031,
+      "learning_rate": 4.3724e-07,
+      "loss": 0.4423,
+      "step": 140750
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 86.27606964111328,
+      "learning_rate": 4.3703999999999995e-07,
+      "loss": 0.3281,
+      "step": 140800
+    },
+    {
+      "epoch": 1.4085,
+      "grad_norm": 0.331612229347229,
+      "learning_rate": 4.3684e-07,
+      "loss": 0.552,
+      "step": 140850
+    },
+    {
+      "epoch": 1.409,
+      "grad_norm": 17.078815460205078,
+      "learning_rate": 4.3663999999999997e-07,
+      "loss": 0.4963,
+      "step": 140900
+    },
+    {
+      "epoch": 1.4095,
+      "grad_norm": 17.343914031982422,
+      "learning_rate": 4.3643999999999996e-07,
+      "loss": 0.3181,
+      "step": 140950
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 75.39768981933594,
+      "learning_rate": 4.3624e-07,
+      "loss": 0.3793,
+      "step": 141000
+    },
+    {
+      "epoch": 1.4104999999999999,
+      "grad_norm": 0.2830001711845398,
+      "learning_rate": 4.3604e-07,
+      "loss": 0.4037,
+      "step": 141050
+    },
+    {
+      "epoch": 1.411,
+      "grad_norm": 30.401926040649414,
+      "learning_rate": 4.3583999999999996e-07,
+      "loss": 0.427,
+      "step": 141100
+    },
+    {
+      "epoch": 1.4115,
+      "grad_norm": 10.625136375427246,
+      "learning_rate": 4.3564e-07,
+      "loss": 0.4488,
+      "step": 141150
+    },
+    {
+      "epoch": 1.412,
+      "grad_norm": 8.427471160888672,
+      "learning_rate": 4.3544e-07,
+      "loss": 0.3985,
+      "step": 141200
+    },
+    {
+      "epoch": 1.4125,
+      "grad_norm": 26.40860366821289,
+      "learning_rate": 4.3523999999999997e-07,
+      "loss": 0.3947,
+      "step": 141250
+    },
+    {
+      "epoch": 1.413,
+      "grad_norm": 87.355224609375,
+      "learning_rate": 4.3503999999999996e-07,
+      "loss": 0.3551,
+      "step": 141300
+    },
+    {
+      "epoch": 1.4135,
+      "grad_norm": 31.92610740661621,
+      "learning_rate": 4.3484e-07,
+      "loss": 0.5354,
+      "step": 141350
+    },
+    {
+      "epoch": 1.414,
+      "grad_norm": 24.853439331054688,
+      "learning_rate": 4.3464000000000003e-07,
+      "loss": 0.4651,
+      "step": 141400
+    },
+    {
+      "epoch": 1.4144999999999999,
+      "grad_norm": 73.6370849609375,
+      "learning_rate": 4.3443999999999996e-07,
+      "loss": 0.4862,
+      "step": 141450
+    },
+    {
+      "epoch": 1.415,
+      "grad_norm": 35.83401870727539,
+      "learning_rate": 4.3424e-07,
+      "loss": 0.4366,
+      "step": 141500
+    },
+    {
+      "epoch": 1.4155,
+      "grad_norm": 30.853010177612305,
+      "learning_rate": 4.3404e-07,
+      "loss": 0.5262,
+      "step": 141550
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 118.53955078125,
+      "learning_rate": 4.3383999999999997e-07,
+      "loss": 0.5409,
+      "step": 141600
+    },
+    {
+      "epoch": 1.4165,
+      "grad_norm": 40.07814025878906,
+      "learning_rate": 4.3364399999999996e-07,
+      "loss": 0.4521,
+      "step": 141650
+    },
+    {
+      "epoch": 1.417,
+      "grad_norm": 9.961045265197754,
+      "learning_rate": 4.33444e-07,
+      "loss": 0.3655,
+      "step": 141700
+    },
+    {
+      "epoch": 1.4175,
+      "grad_norm": 141.18646240234375,
+      "learning_rate": 4.33244e-07,
+      "loss": 0.6208,
+      "step": 141750
+    },
+    {
+      "epoch": 1.418,
+      "grad_norm": 58.834197998046875,
+      "learning_rate": 4.3304399999999997e-07,
+      "loss": 0.3911,
+      "step": 141800
+    },
+    {
+      "epoch": 1.4184999999999999,
+      "grad_norm": 63.38532257080078,
+      "learning_rate": 4.32844e-07,
+      "loss": 0.3999,
+      "step": 141850
+    },
+    {
+      "epoch": 1.419,
+      "grad_norm": 77.08700561523438,
+      "learning_rate": 4.3264399999999994e-07,
+      "loss": 0.5151,
+      "step": 141900
+    },
+    {
+      "epoch": 1.4195,
+      "grad_norm": 0.9496738910675049,
+      "learning_rate": 4.32444e-07,
+      "loss": 0.4697,
+      "step": 141950
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 35.80956268310547,
+      "learning_rate": 4.32244e-07,
+      "loss": 0.3883,
+      "step": 142000
+    },
+    {
+      "epoch": 1.4205,
+      "grad_norm": 82.25574493408203,
+      "learning_rate": 4.3204399999999995e-07,
+      "loss": 0.5685,
+      "step": 142050
+    },
+    {
+      "epoch": 1.421,
+      "grad_norm": 8.83198356628418,
+      "learning_rate": 4.31844e-07,
+      "loss": 0.4466,
+      "step": 142100
+    },
+    {
+      "epoch": 1.4215,
+      "grad_norm": 84.17395782470703,
+      "learning_rate": 4.31644e-07,
+      "loss": 0.4718,
+      "step": 142150
+    },
+    {
+      "epoch": 1.422,
+      "grad_norm": 94.86622619628906,
+      "learning_rate": 4.3144399999999996e-07,
+      "loss": 0.3978,
+      "step": 142200
+    },
+    {
+      "epoch": 1.4224999999999999,
+      "grad_norm": 77.70457458496094,
+      "learning_rate": 4.31244e-07,
+      "loss": 0.4577,
+      "step": 142250
+    },
+    {
+      "epoch": 1.423,
+      "grad_norm": 95.18637084960938,
+      "learning_rate": 4.31044e-07,
+      "loss": 0.4149,
+      "step": 142300
+    },
+    {
+      "epoch": 1.4235,
+      "grad_norm": 63.49544143676758,
+      "learning_rate": 4.3084399999999996e-07,
+      "loss": 0.5862,
+      "step": 142350
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 16.329381942749023,
+      "learning_rate": 4.30644e-07,
+      "loss": 0.3519,
+      "step": 142400
+    },
+    {
+      "epoch": 1.4245,
+      "grad_norm": 8.99925708770752,
+      "learning_rate": 4.30444e-07,
+      "loss": 0.3929,
+      "step": 142450
+    },
+    {
+      "epoch": 1.425,
+      "grad_norm": 3.1051368713378906,
+      "learning_rate": 4.3024399999999997e-07,
+      "loss": 0.3946,
+      "step": 142500
+    },
+    {
+      "epoch": 1.4255,
+      "grad_norm": 116.58087921142578,
+      "learning_rate": 4.3004399999999996e-07,
+      "loss": 0.5392,
+      "step": 142550
+    },
+    {
+      "epoch": 1.426,
+      "grad_norm": 11.166736602783203,
+      "learning_rate": 4.29848e-07,
+      "loss": 0.4117,
+      "step": 142600
+    },
+    {
+      "epoch": 1.4264999999999999,
+      "grad_norm": 45.40515899658203,
+      "learning_rate": 4.2964799999999993e-07,
+      "loss": 0.4275,
+      "step": 142650
+    },
+    {
+      "epoch": 1.427,
+      "grad_norm": 49.694183349609375,
+      "learning_rate": 4.2944799999999997e-07,
+      "loss": 0.3936,
+      "step": 142700
+    },
+    {
+      "epoch": 1.4275,
+      "grad_norm": 52.98374557495117,
+      "learning_rate": 4.29248e-07,
+      "loss": 0.4006,
+      "step": 142750
+    },
+    {
+      "epoch": 1.428,
+      "grad_norm": 23.433656692504883,
+      "learning_rate": 4.29048e-07,
+      "loss": 0.4007,
+      "step": 142800
+    },
+    {
+      "epoch": 1.4285,
+      "grad_norm": 88.89615631103516,
+      "learning_rate": 4.28848e-07,
+      "loss": 0.442,
+      "step": 142850
+    },
+    {
+      "epoch": 1.429,
+      "grad_norm": 2.4820308685302734,
+      "learning_rate": 4.2864799999999996e-07,
+      "loss": 0.4532,
+      "step": 142900
+    },
+    {
+      "epoch": 1.4295,
+      "grad_norm": 26.804426193237305,
+      "learning_rate": 4.28448e-07,
+      "loss": 0.4837,
+      "step": 142950
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.690469741821289,
+      "learning_rate": 4.28248e-07,
+      "loss": 0.4484,
+      "step": 143000
+    },
+    {
+      "epoch": 1.4304999999999999,
+      "grad_norm": 31.82974624633789,
+      "learning_rate": 4.2804799999999997e-07,
+      "loss": 0.4718,
+      "step": 143050
+    },
+    {
+      "epoch": 1.431,
+      "grad_norm": 76.13028717041016,
+      "learning_rate": 4.27848e-07,
+      "loss": 0.4575,
+      "step": 143100
+    },
+    {
+      "epoch": 1.4315,
+      "grad_norm": 79.27205657958984,
+      "learning_rate": 4.2764799999999994e-07,
+      "loss": 0.4532,
+      "step": 143150
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 1.6054680347442627,
+      "learning_rate": 4.27448e-07,
+      "loss": 0.4233,
+      "step": 143200
+    },
+    {
+      "epoch": 1.4325,
+      "grad_norm": 5.973708629608154,
+      "learning_rate": 4.27248e-07,
+      "loss": 0.4445,
+      "step": 143250
+    },
+    {
+      "epoch": 1.433,
+      "grad_norm": 97.4744873046875,
+      "learning_rate": 4.2704799999999995e-07,
+      "loss": 0.5598,
+      "step": 143300
+    },
+    {
+      "epoch": 1.4335,
+      "grad_norm": 7.280982494354248,
+      "learning_rate": 4.26848e-07,
+      "loss": 0.4821,
+      "step": 143350
+    },
+    {
+      "epoch": 1.434,
+      "grad_norm": 30.4831485748291,
+      "learning_rate": 4.26648e-07,
+      "loss": 0.3772,
+      "step": 143400
+    },
+    {
+      "epoch": 1.4344999999999999,
+      "grad_norm": 6.623522758483887,
+      "learning_rate": 4.2644799999999996e-07,
+      "loss": 0.4908,
+      "step": 143450
+    },
+    {
+      "epoch": 1.435,
+      "grad_norm": 26.69012451171875,
+      "learning_rate": 4.26248e-07,
+      "loss": 0.4367,
+      "step": 143500
+    },
+    {
+      "epoch": 1.4355,
+      "grad_norm": 48.03159713745117,
+      "learning_rate": 4.26048e-07,
+      "loss": 0.57,
+      "step": 143550
+    },
+    {
+      "epoch": 1.436,
+      "grad_norm": 0.36997270584106445,
+      "learning_rate": 4.2584799999999996e-07,
+      "loss": 0.381,
+      "step": 143600
+    },
+    {
+      "epoch": 1.4365,
+      "grad_norm": 75.4528579711914,
+      "learning_rate": 4.25648e-07,
+      "loss": 0.3569,
+      "step": 143650
+    },
+    {
+      "epoch": 1.437,
+      "grad_norm": 23.47142219543457,
+      "learning_rate": 4.25448e-07,
+      "loss": 0.4182,
+      "step": 143700
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 40.28401184082031,
+      "learning_rate": 4.2524799999999997e-07,
+      "loss": 0.4932,
+      "step": 143750
+    },
+    {
+      "epoch": 1.438,
+      "grad_norm": 8.089696884155273,
+      "learning_rate": 4.2504799999999996e-07,
+      "loss": 0.5582,
+      "step": 143800
+    },
+    {
+      "epoch": 1.4385,
+      "grad_norm": 93.84317779541016,
+      "learning_rate": 4.24848e-07,
+      "loss": 0.3421,
+      "step": 143850
+    },
+    {
+      "epoch": 1.439,
+      "grad_norm": 78.79322052001953,
+      "learning_rate": 4.2464800000000003e-07,
+      "loss": 0.5463,
+      "step": 143900
+    },
+    {
+      "epoch": 1.4395,
+      "grad_norm": 153.0576171875,
+      "learning_rate": 4.2444799999999996e-07,
+      "loss": 0.4011,
+      "step": 143950
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.934996604919434,
+      "learning_rate": 4.24248e-07,
+      "loss": 0.4774,
+      "step": 144000
+    },
+    {
+      "epoch": 1.4405000000000001,
+      "grad_norm": 86.63742065429688,
+      "learning_rate": 4.24048e-07,
+      "loss": 0.4094,
+      "step": 144050
+    },
+    {
+      "epoch": 1.441,
+      "grad_norm": 79.47491455078125,
+      "learning_rate": 4.2384799999999997e-07,
+      "loss": 0.3831,
+      "step": 144100
+    },
+    {
+      "epoch": 1.4415,
+      "grad_norm": 55.07907485961914,
+      "learning_rate": 4.23648e-07,
+      "loss": 0.4505,
+      "step": 144150
+    },
+    {
+      "epoch": 1.442,
+      "grad_norm": 7.138432025909424,
+      "learning_rate": 4.23448e-07,
+      "loss": 0.4995,
+      "step": 144200
+    },
+    {
+      "epoch": 1.4425,
+      "grad_norm": 31.88157844543457,
+      "learning_rate": 4.23248e-07,
+      "loss": 0.674,
+      "step": 144250
+    },
+    {
+      "epoch": 1.443,
+      "grad_norm": 24.490238189697266,
+      "learning_rate": 4.2304799999999996e-07,
+      "loss": 0.4163,
+      "step": 144300
+    },
+    {
+      "epoch": 1.4435,
+      "grad_norm": 67.5228042602539,
+      "learning_rate": 4.22848e-07,
+      "loss": 0.3403,
+      "step": 144350
+    },
+    {
+      "epoch": 1.444,
+      "grad_norm": 1.8344687223434448,
+      "learning_rate": 4.22648e-07,
+      "loss": 0.5146,
+      "step": 144400
+    },
+    {
+      "epoch": 1.4445000000000001,
+      "grad_norm": 52.78595733642578,
+      "learning_rate": 4.2244799999999997e-07,
+      "loss": 0.5185,
+      "step": 144450
+    },
+    {
+      "epoch": 1.445,
+      "grad_norm": 8.644986152648926,
+      "learning_rate": 4.22248e-07,
+      "loss": 0.5884,
+      "step": 144500
+    },
+    {
+      "epoch": 1.4455,
+      "grad_norm": 28.123138427734375,
+      "learning_rate": 4.2204799999999994e-07,
+      "loss": 0.4497,
+      "step": 144550
+    },
+    {
+      "epoch": 1.446,
+      "grad_norm": 67.64913177490234,
+      "learning_rate": 4.21848e-07,
+      "loss": 0.4375,
+      "step": 144600
+    },
+    {
+      "epoch": 1.4465,
+      "grad_norm": 30.597097396850586,
+      "learning_rate": 4.21648e-07,
+      "loss": 0.5279,
+      "step": 144650
+    },
+    {
+      "epoch": 1.447,
+      "grad_norm": 100.2088851928711,
+      "learning_rate": 4.2144799999999995e-07,
+      "loss": 0.5009,
+      "step": 144700
+    },
+    {
+      "epoch": 1.4475,
+      "grad_norm": 55.302974700927734,
+      "learning_rate": 4.21248e-07,
+      "loss": 0.455,
+      "step": 144750
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 67.07560729980469,
+      "learning_rate": 4.2104799999999997e-07,
+      "loss": 0.4658,
+      "step": 144800
+    },
+    {
+      "epoch": 1.4485000000000001,
+      "grad_norm": 13.170785903930664,
+      "learning_rate": 4.2084799999999996e-07,
+      "loss": 0.3561,
+      "step": 144850
+    },
+    {
+      "epoch": 1.449,
+      "grad_norm": 7.08843994140625,
+      "learning_rate": 4.20648e-07,
+      "loss": 0.3366,
+      "step": 144900
+    },
+    {
+      "epoch": 1.4495,
+      "grad_norm": 3.2631988525390625,
+      "learning_rate": 4.20448e-07,
+      "loss": 0.435,
+      "step": 144950
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.511624336242676,
+      "learning_rate": 4.2024799999999997e-07,
+      "loss": 0.2825,
+      "step": 145000
+    },
+    {
+      "epoch": 1.4505,
+      "grad_norm": 13.6480073928833,
+      "learning_rate": 4.2004799999999995e-07,
+      "loss": 0.4737,
+      "step": 145050
+    },
+    {
+      "epoch": 1.451,
+      "grad_norm": 84.39595794677734,
+      "learning_rate": 4.19848e-07,
+      "loss": 0.478,
+      "step": 145100
+    },
+    {
+      "epoch": 1.4515,
+      "grad_norm": 112.03206634521484,
+      "learning_rate": 4.1964800000000003e-07,
+      "loss": 0.3953,
+      "step": 145150
+    },
+    {
+      "epoch": 1.452,
+      "grad_norm": 13.859879493713379,
+      "learning_rate": 4.1944799999999996e-07,
+      "loss": 0.4724,
+      "step": 145200
+    },
+    {
+      "epoch": 1.4525000000000001,
+      "grad_norm": 51.57634735107422,
+      "learning_rate": 4.19248e-07,
+      "loss": 0.4693,
+      "step": 145250
+    },
+    {
+      "epoch": 1.453,
+      "grad_norm": 1.5918205976486206,
+      "learning_rate": 4.19048e-07,
+      "loss": 0.3019,
+      "step": 145300
+    },
+    {
+      "epoch": 1.4535,
+      "grad_norm": 86.82250213623047,
+      "learning_rate": 4.1884799999999997e-07,
+      "loss": 0.4294,
+      "step": 145350
+    },
+    {
+      "epoch": 1.454,
+      "grad_norm": 29.19518280029297,
+      "learning_rate": 4.18648e-07,
+      "loss": 0.4965,
+      "step": 145400
+    },
+    {
+      "epoch": 1.4545,
+      "grad_norm": 38.55145263671875,
+      "learning_rate": 4.18448e-07,
+      "loss": 0.3507,
+      "step": 145450
+    },
+    {
+      "epoch": 1.455,
+      "grad_norm": 96.74462127685547,
+      "learning_rate": 4.18248e-07,
+      "loss": 0.5309,
+      "step": 145500
+    },
+    {
+      "epoch": 1.4555,
+      "grad_norm": 71.57083129882812,
+      "learning_rate": 4.1804799999999996e-07,
+      "loss": 0.5173,
+      "step": 145550
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 142.42713928222656,
+      "learning_rate": 4.17848e-07,
+      "loss": 0.4896,
+      "step": 145600
+    },
+    {
+      "epoch": 1.4565000000000001,
+      "grad_norm": 71.31710815429688,
+      "learning_rate": 4.17648e-07,
+      "loss": 0.4368,
+      "step": 145650
+    },
+    {
+      "epoch": 1.457,
+      "grad_norm": 40.85289001464844,
+      "learning_rate": 4.1744799999999997e-07,
+      "loss": 0.4845,
+      "step": 145700
+    },
+    {
+      "epoch": 1.4575,
+      "grad_norm": 92.76287841796875,
+      "learning_rate": 4.17248e-07,
+      "loss": 0.4203,
+      "step": 145750
+    },
+    {
+      "epoch": 1.458,
+      "grad_norm": 34.11021423339844,
+      "learning_rate": 4.1704799999999994e-07,
+      "loss": 0.3603,
+      "step": 145800
+    },
+    {
+      "epoch": 1.4585,
+      "grad_norm": 19.930692672729492,
+      "learning_rate": 4.16848e-07,
+      "loss": 0.4356,
+      "step": 145850
+    },
+    {
+      "epoch": 1.459,
+      "grad_norm": 9.085647583007812,
+      "learning_rate": 4.16648e-07,
+      "loss": 0.3679,
+      "step": 145900
+    },
+    {
+      "epoch": 1.4595,
+      "grad_norm": 127.162841796875,
+      "learning_rate": 4.1644799999999994e-07,
+      "loss": 0.4983,
+      "step": 145950
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 26.87095069885254,
+      "learning_rate": 4.16248e-07,
+      "loss": 0.452,
+      "step": 146000
+    },
+    {
+      "epoch": 1.4605000000000001,
+      "grad_norm": 6.62654972076416,
+      "learning_rate": 4.1604799999999997e-07,
+      "loss": 0.4967,
+      "step": 146050
+    },
+    {
+      "epoch": 1.461,
+      "grad_norm": 0.27412235736846924,
+      "learning_rate": 4.1584799999999995e-07,
+      "loss": 0.35,
+      "step": 146100
+    },
+    {
+      "epoch": 1.4615,
+      "grad_norm": 0.39492130279541016,
+      "learning_rate": 4.15648e-07,
+      "loss": 0.5392,
+      "step": 146150
+    },
+    {
+      "epoch": 1.462,
+      "grad_norm": 0.3583661615848541,
+      "learning_rate": 4.15448e-07,
+      "loss": 0.4021,
+      "step": 146200
+    },
+    {
+      "epoch": 1.4625,
+      "grad_norm": 66.29902648925781,
+      "learning_rate": 4.15248e-07,
+      "loss": 0.4798,
+      "step": 146250
+    },
+    {
+      "epoch": 1.463,
+      "grad_norm": 14.90131664276123,
+      "learning_rate": 4.1504799999999995e-07,
+      "loss": 0.4671,
+      "step": 146300
+    },
+    {
+      "epoch": 1.4635,
+      "grad_norm": 8.83768367767334,
+      "learning_rate": 4.14848e-07,
+      "loss": 0.5414,
+      "step": 146350
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 39.12753677368164,
+      "learning_rate": 4.14648e-07,
+      "loss": 0.5358,
+      "step": 146400
+    },
+    {
+      "epoch": 1.4645000000000001,
+      "grad_norm": 50.82841110229492,
+      "learning_rate": 4.1444799999999995e-07,
+      "loss": 0.4129,
+      "step": 146450
+    },
+    {
+      "epoch": 1.465,
+      "grad_norm": 39.67839431762695,
+      "learning_rate": 4.14248e-07,
+      "loss": 0.4745,
+      "step": 146500
+    },
+    {
+      "epoch": 1.4655,
+      "grad_norm": 89.63794708251953,
+      "learning_rate": 4.1404800000000003e-07,
+      "loss": 0.5237,
+      "step": 146550
+    },
+    {
+      "epoch": 1.466,
+      "grad_norm": 65.97050476074219,
+      "learning_rate": 4.1384799999999996e-07,
+      "loss": 0.4113,
+      "step": 146600
+    },
+    {
+      "epoch": 1.4665,
+      "grad_norm": 30.679176330566406,
+      "learning_rate": 4.13648e-07,
+      "loss": 0.5226,
+      "step": 146650
+    },
+    {
+      "epoch": 1.467,
+      "grad_norm": 75.3191909790039,
+      "learning_rate": 4.13448e-07,
+      "loss": 0.4449,
+      "step": 146700
+    },
+    {
+      "epoch": 1.4675,
+      "grad_norm": 63.889930725097656,
+      "learning_rate": 4.1324799999999997e-07,
+      "loss": 0.4736,
+      "step": 146750
+    },
+    {
+      "epoch": 1.468,
+      "grad_norm": 61.48563003540039,
+      "learning_rate": 4.13048e-07,
+      "loss": 0.3692,
+      "step": 146800
+    },
+    {
+      "epoch": 1.4685000000000001,
+      "grad_norm": 34.46501541137695,
+      "learning_rate": 4.12848e-07,
+      "loss": 0.3301,
+      "step": 146850
+    },
+    {
+      "epoch": 1.4689999999999999,
+      "grad_norm": 11.682842254638672,
+      "learning_rate": 4.12648e-07,
+      "loss": 0.4875,
+      "step": 146900
+    },
+    {
+      "epoch": 1.4695,
+      "grad_norm": 6.080808639526367,
+      "learning_rate": 4.1244799999999996e-07,
+      "loss": 0.5323,
+      "step": 146950
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 29.79290008544922,
+      "learning_rate": 4.12248e-07,
+      "loss": 0.5893,
+      "step": 147000
+    },
+    {
+      "epoch": 1.4705,
+      "grad_norm": 41.09132766723633,
+      "learning_rate": 4.12048e-07,
+      "loss": 0.4248,
+      "step": 147050
+    },
+    {
+      "epoch": 1.471,
+      "grad_norm": 62.73756408691406,
+      "learning_rate": 4.1184799999999997e-07,
+      "loss": 0.4393,
+      "step": 147100
+    },
+    {
+      "epoch": 1.4715,
+      "grad_norm": 41.83452224731445,
+      "learning_rate": 4.11648e-07,
+      "loss": 0.396,
+      "step": 147150
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 59.541011810302734,
+      "learning_rate": 4.1144799999999994e-07,
+      "loss": 0.4265,
+      "step": 147200
+    },
+    {
+      "epoch": 1.4725,
+      "grad_norm": 23.928707122802734,
+      "learning_rate": 4.11248e-07,
+      "loss": 0.3726,
+      "step": 147250
+    },
+    {
+      "epoch": 1.4729999999999999,
+      "grad_norm": 70.83897399902344,
+      "learning_rate": 4.11048e-07,
+      "loss": 0.3918,
+      "step": 147300
+    },
+    {
+      "epoch": 1.4735,
+      "grad_norm": 21.586318969726562,
+      "learning_rate": 4.1084799999999995e-07,
+      "loss": 0.3456,
+      "step": 147350
+    },
+    {
+      "epoch": 1.474,
+      "grad_norm": 11.594403266906738,
+      "learning_rate": 4.10648e-07,
+      "loss": 0.3408,
+      "step": 147400
+    },
+    {
+      "epoch": 1.4745,
+      "grad_norm": 47.519081115722656,
+      "learning_rate": 4.1044799999999997e-07,
+      "loss": 0.3752,
+      "step": 147450
+    },
+    {
+      "epoch": 1.475,
+      "grad_norm": 128.55673217773438,
+      "learning_rate": 4.10248e-07,
+      "loss": 0.4493,
+      "step": 147500
+    },
+    {
+      "epoch": 1.4755,
+      "grad_norm": 67.64390563964844,
+      "learning_rate": 4.10048e-07,
+      "loss": 0.5414,
+      "step": 147550
+    },
+    {
+      "epoch": 1.476,
+      "grad_norm": 35.612281799316406,
+      "learning_rate": 4.09848e-07,
+      "loss": 0.4347,
+      "step": 147600
+    },
+    {
+      "epoch": 1.4765,
+      "grad_norm": 71.22669219970703,
+      "learning_rate": 4.09648e-07,
+      "loss": 0.4211,
+      "step": 147650
+    },
+    {
+      "epoch": 1.4769999999999999,
+      "grad_norm": 79.02700805664062,
+      "learning_rate": 4.0944799999999995e-07,
+      "loss": 0.3902,
+      "step": 147700
+    },
+    {
+      "epoch": 1.4775,
+      "grad_norm": 118.43605041503906,
+      "learning_rate": 4.09248e-07,
+      "loss": 0.4709,
+      "step": 147750
+    },
+    {
+      "epoch": 1.478,
+      "grad_norm": 42.166351318359375,
+      "learning_rate": 4.09048e-07,
+      "loss": 0.2732,
+      "step": 147800
+    },
+    {
+      "epoch": 1.4785,
+      "grad_norm": 12.510908126831055,
+      "learning_rate": 4.0884799999999996e-07,
+      "loss": 0.3895,
+      "step": 147850
+    },
+    {
+      "epoch": 1.479,
+      "grad_norm": 91.17105102539062,
+      "learning_rate": 4.08648e-07,
+      "loss": 0.4198,
+      "step": 147900
+    },
+    {
+      "epoch": 1.4795,
+      "grad_norm": 69.4258041381836,
+      "learning_rate": 4.08448e-07,
+      "loss": 0.4962,
+      "step": 147950
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 45.617069244384766,
+      "learning_rate": 4.0824799999999996e-07,
+      "loss": 0.3913,
+      "step": 148000
+    },
+    {
+      "epoch": 1.4805,
+      "grad_norm": 88.43484497070312,
+      "learning_rate": 4.08048e-07,
+      "loss": 0.3713,
+      "step": 148050
+    },
+    {
+      "epoch": 1.4809999999999999,
+      "grad_norm": 28.707489013671875,
+      "learning_rate": 4.07848e-07,
+      "loss": 0.377,
+      "step": 148100
+    },
+    {
+      "epoch": 1.4815,
+      "grad_norm": 29.1148681640625,
+      "learning_rate": 4.0764799999999997e-07,
+      "loss": 0.3606,
+      "step": 148150
+    },
+    {
+      "epoch": 1.482,
+      "grad_norm": 11.052894592285156,
+      "learning_rate": 4.0744799999999996e-07,
+      "loss": 0.4278,
+      "step": 148200
+    },
+    {
+      "epoch": 1.4825,
+      "grad_norm": 94.74927520751953,
+      "learning_rate": 4.07248e-07,
+      "loss": 0.3981,
+      "step": 148250
+    },
+    {
+      "epoch": 1.483,
+      "grad_norm": 28.74058723449707,
+      "learning_rate": 4.07048e-07,
+      "loss": 0.5204,
+      "step": 148300
+    },
+    {
+      "epoch": 1.4835,
+      "grad_norm": 44.29704284667969,
+      "learning_rate": 4.0684799999999996e-07,
+      "loss": 0.473,
+      "step": 148350
+    },
+    {
+      "epoch": 1.484,
+      "grad_norm": 3.45829439163208,
+      "learning_rate": 4.06648e-07,
+      "loss": 0.423,
+      "step": 148400
+    },
+    {
+      "epoch": 1.4845,
+      "grad_norm": 112.15509796142578,
+      "learning_rate": 4.0644799999999993e-07,
+      "loss": 0.4297,
+      "step": 148450
+    },
+    {
+      "epoch": 1.4849999999999999,
+      "grad_norm": 37.767704010009766,
+      "learning_rate": 4.0624799999999997e-07,
+      "loss": 0.407,
+      "step": 148500
+    },
+    {
+      "epoch": 1.4855,
+      "grad_norm": 114.10420227050781,
+      "learning_rate": 4.06048e-07,
+      "loss": 0.49,
+      "step": 148550
+    },
+    {
+      "epoch": 1.486,
+      "grad_norm": 81.723388671875,
+      "learning_rate": 4.0584799999999994e-07,
+      "loss": 0.3653,
+      "step": 148600
+    },
+    {
+      "epoch": 1.4865,
+      "grad_norm": 2.658217191696167,
+      "learning_rate": 4.05652e-07,
+      "loss": 0.4765,
+      "step": 148650
+    },
+    {
+      "epoch": 1.487,
+      "grad_norm": 0.6626757979393005,
+      "learning_rate": 4.0545199999999997e-07,
+      "loss": 0.4804,
+      "step": 148700
+    },
+    {
+      "epoch": 1.4875,
+      "grad_norm": 71.02579498291016,
+      "learning_rate": 4.05252e-07,
+      "loss": 0.4608,
+      "step": 148750
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 7.002717018127441,
+      "learning_rate": 4.05052e-07,
+      "loss": 0.4567,
+      "step": 148800
+    },
+    {
+      "epoch": 1.4885,
+      "grad_norm": 159.69973754882812,
+      "learning_rate": 4.04852e-07,
+      "loss": 0.4012,
+      "step": 148850
+    },
+    {
+      "epoch": 1.4889999999999999,
+      "grad_norm": 33.81310272216797,
+      "learning_rate": 4.04652e-07,
+      "loss": 0.4608,
+      "step": 148900
+    },
+    {
+      "epoch": 1.4895,
+      "grad_norm": 79.68433380126953,
+      "learning_rate": 4.0445199999999995e-07,
+      "loss": 0.4923,
+      "step": 148950
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.4748802185058594,
+      "learning_rate": 4.04252e-07,
+      "loss": 0.39,
+      "step": 149000
+    },
+    {
+      "epoch": 1.4905,
+      "grad_norm": 53.5771598815918,
+      "learning_rate": 4.04052e-07,
+      "loss": 0.5652,
+      "step": 149050
+    },
+    {
+      "epoch": 1.491,
+      "grad_norm": 46.88816452026367,
+      "learning_rate": 4.0385199999999996e-07,
+      "loss": 0.4258,
+      "step": 149100
+    },
+    {
+      "epoch": 1.4915,
+      "grad_norm": 32.95225524902344,
+      "learning_rate": 4.03652e-07,
+      "loss": 0.4477,
+      "step": 149150
+    },
+    {
+      "epoch": 1.492,
+      "grad_norm": 68.19593048095703,
+      "learning_rate": 4.03452e-07,
+      "loss": 0.4861,
+      "step": 149200
+    },
+    {
+      "epoch": 1.4925,
+      "grad_norm": 78.71890258789062,
+      "learning_rate": 4.0325199999999996e-07,
+      "loss": 0.5034,
+      "step": 149250
+    },
+    {
+      "epoch": 1.4929999999999999,
+      "grad_norm": 4.197335720062256,
+      "learning_rate": 4.03052e-07,
+      "loss": 0.5138,
+      "step": 149300
+    },
+    {
+      "epoch": 1.4935,
+      "grad_norm": 6.181484222412109,
+      "learning_rate": 4.02852e-07,
+      "loss": 0.4473,
+      "step": 149350
+    },
+    {
+      "epoch": 1.494,
+      "grad_norm": 5.397106170654297,
+      "learning_rate": 4.0265199999999997e-07,
+      "loss": 0.4057,
+      "step": 149400
+    },
+    {
+      "epoch": 1.4945,
+      "grad_norm": 6.172823905944824,
+      "learning_rate": 4.0245199999999996e-07,
+      "loss": 0.4547,
+      "step": 149450
+    },
+    {
+      "epoch": 1.495,
+      "grad_norm": 30.402366638183594,
+      "learning_rate": 4.02252e-07,
+      "loss": 0.4936,
+      "step": 149500
+    },
+    {
+      "epoch": 1.4955,
+      "grad_norm": 90.25249481201172,
+      "learning_rate": 4.02052e-07,
+      "loss": 0.4147,
+      "step": 149550
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 5.896697521209717,
+      "learning_rate": 4.0185199999999996e-07,
+      "loss": 0.4432,
+      "step": 149600
+    },
+    {
+      "epoch": 1.4965,
+      "grad_norm": 66.12384796142578,
+      "learning_rate": 4.01652e-07,
+      "loss": 0.4179,
+      "step": 149650
+    },
+    {
+      "epoch": 1.4969999999999999,
+      "grad_norm": 26.684412002563477,
+      "learning_rate": 4.0145199999999993e-07,
+      "loss": 0.5063,
+      "step": 149700
+    },
+    {
+      "epoch": 1.4975,
+      "grad_norm": 0.9913628697395325,
+      "learning_rate": 4.0125199999999997e-07,
+      "loss": 0.3879,
+      "step": 149750
+    },
+    {
+      "epoch": 1.498,
+      "grad_norm": 98.07878875732422,
+      "learning_rate": 4.01052e-07,
+      "loss": 0.4664,
+      "step": 149800
+    },
+    {
+      "epoch": 1.4985,
+      "grad_norm": 78.66146850585938,
+      "learning_rate": 4.0085199999999994e-07,
+      "loss": 0.5213,
+      "step": 149850
+    },
+    {
+      "epoch": 1.499,
+      "grad_norm": 18.173500061035156,
+      "learning_rate": 4.00652e-07,
+      "loss": 0.3256,
+      "step": 149900
+    },
+    {
+      "epoch": 1.4995,
+      "grad_norm": 0.6916254758834839,
+      "learning_rate": 4.0045199999999997e-07,
+      "loss": 0.4523,
+      "step": 149950
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.336428642272949,
+      "learning_rate": 4.00252e-07,
+      "loss": 0.3649,
+      "step": 150000
+    },
+    {
+      "epoch": 1.5005,
+      "grad_norm": 72.81371307373047,
+      "learning_rate": 4.00052e-07,
+      "loss": 0.4134,
+      "step": 150050
+    },
+    {
+      "epoch": 1.501,
+      "grad_norm": 53.876487731933594,
+      "learning_rate": 3.9985199999999997e-07,
+      "loss": 0.4047,
+      "step": 150100
+    },
+    {
+      "epoch": 1.5015,
+      "grad_norm": 146.40159606933594,
+      "learning_rate": 3.99652e-07,
+      "loss": 0.4675,
+      "step": 150150
+    },
+    {
+      "epoch": 1.502,
+      "grad_norm": 65.77076721191406,
+      "learning_rate": 3.9945199999999994e-07,
+      "loss": 0.5416,
+      "step": 150200
+    },
+    {
+      "epoch": 1.5025,
+      "grad_norm": 41.62353515625,
+      "learning_rate": 3.99252e-07,
+      "loss": 0.5011,
+      "step": 150250
+    },
+    {
+      "epoch": 1.5030000000000001,
+      "grad_norm": 75.19416809082031,
+      "learning_rate": 3.99052e-07,
+      "loss": 0.3859,
+      "step": 150300
+    },
+    {
+      "epoch": 1.5034999999999998,
+      "grad_norm": 26.303030014038086,
+      "learning_rate": 3.9885199999999995e-07,
+      "loss": 0.4655,
+      "step": 150350
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 97.27670288085938,
+      "learning_rate": 3.98652e-07,
+      "loss": 0.4718,
+      "step": 150400
+    },
+    {
+      "epoch": 1.5045,
+      "grad_norm": 120.45020294189453,
+      "learning_rate": 3.98456e-07,
+      "loss": 0.402,
+      "step": 150450
+    },
+    {
+      "epoch": 1.505,
+      "grad_norm": 57.528079986572266,
+      "learning_rate": 3.9825599999999996e-07,
+      "loss": 0.4117,
+      "step": 150500
+    },
+    {
+      "epoch": 1.5055,
+      "grad_norm": 110.59357452392578,
+      "learning_rate": 3.98056e-07,
+      "loss": 0.4865,
+      "step": 150550
+    },
+    {
+      "epoch": 1.506,
+      "grad_norm": 20.42061996459961,
+      "learning_rate": 3.97856e-07,
+      "loss": 0.5312,
+      "step": 150600
+    },
+    {
+      "epoch": 1.5065,
+      "grad_norm": 19.152402877807617,
+      "learning_rate": 3.9765599999999997e-07,
+      "loss": 0.5577,
+      "step": 150650
+    },
+    {
+      "epoch": 1.5070000000000001,
+      "grad_norm": 144.61178588867188,
+      "learning_rate": 3.9745599999999996e-07,
+      "loss": 0.4414,
+      "step": 150700
+    },
+    {
+      "epoch": 1.5074999999999998,
+      "grad_norm": 29.284423828125,
+      "learning_rate": 3.97256e-07,
+      "loss": 0.4616,
+      "step": 150750
+    },
+    {
+      "epoch": 1.508,
+      "grad_norm": 5.666861057281494,
+      "learning_rate": 3.97056e-07,
+      "loss": 0.5011,
+      "step": 150800
+    },
+    {
+      "epoch": 1.5085,
+      "grad_norm": 68.09160614013672,
+      "learning_rate": 3.9685599999999996e-07,
+      "loss": 0.5303,
+      "step": 150850
+    },
+    {
+      "epoch": 1.509,
+      "grad_norm": 17.154300689697266,
+      "learning_rate": 3.96656e-07,
+      "loss": 0.5095,
+      "step": 150900
+    },
+    {
+      "epoch": 1.5095,
+      "grad_norm": 76.5134048461914,
+      "learning_rate": 3.9645599999999993e-07,
+      "loss": 0.446,
+      "step": 150950
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.559085845947266,
+      "learning_rate": 3.9625599999999997e-07,
+      "loss": 0.5092,
+      "step": 151000
+    },
+    {
+      "epoch": 1.5105,
+      "grad_norm": 60.698848724365234,
+      "learning_rate": 3.96056e-07,
+      "loss": 0.5773,
+      "step": 151050
+    },
+    {
+      "epoch": 1.5110000000000001,
+      "grad_norm": 122.44525909423828,
+      "learning_rate": 3.9585599999999994e-07,
+      "loss": 0.4852,
+      "step": 151100
+    },
+    {
+      "epoch": 1.5114999999999998,
+      "grad_norm": 31.7929744720459,
+      "learning_rate": 3.95656e-07,
+      "loss": 0.3784,
+      "step": 151150
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 73.54200744628906,
+      "learning_rate": 3.9545599999999996e-07,
+      "loss": 0.3659,
+      "step": 151200
+    },
+    {
+      "epoch": 1.5125,
+      "grad_norm": 45.57496643066406,
+      "learning_rate": 3.95256e-07,
+      "loss": 0.4095,
+      "step": 151250
+    },
+    {
+      "epoch": 1.513,
+      "grad_norm": 35.19536590576172,
+      "learning_rate": 3.95056e-07,
+      "loss": 0.3262,
+      "step": 151300
+    },
+    {
+      "epoch": 1.5135,
+      "grad_norm": 29.406625747680664,
+      "learning_rate": 3.9485599999999997e-07,
+      "loss": 0.2914,
+      "step": 151350
+    },
+    {
+      "epoch": 1.514,
+      "grad_norm": 11.98690128326416,
+      "learning_rate": 3.94656e-07,
+      "loss": 0.5087,
+      "step": 151400
+    },
+    {
+      "epoch": 1.5145,
+      "grad_norm": 404.19842529296875,
+      "learning_rate": 3.9445599999999994e-07,
+      "loss": 0.5389,
+      "step": 151450
+    },
+    {
+      "epoch": 1.5150000000000001,
+      "grad_norm": 71.88636016845703,
+      "learning_rate": 3.94256e-07,
+      "loss": 0.5309,
+      "step": 151500
+    },
+    {
+      "epoch": 1.5154999999999998,
+      "grad_norm": 130.74053955078125,
+      "learning_rate": 3.94056e-07,
+      "loss": 0.4026,
+      "step": 151550
+    },
+    {
+      "epoch": 1.516,
+      "grad_norm": 43.46409606933594,
+      "learning_rate": 3.9385599999999995e-07,
+      "loss": 0.5388,
+      "step": 151600
+    },
+    {
+      "epoch": 1.5165,
+      "grad_norm": 7.446013927459717,
+      "learning_rate": 3.93656e-07,
+      "loss": 0.5019,
+      "step": 151650
+    },
+    {
+      "epoch": 1.517,
+      "grad_norm": 99.18240356445312,
+      "learning_rate": 3.9345599999999997e-07,
+      "loss": 0.5287,
+      "step": 151700
+    },
+    {
+      "epoch": 1.5175,
+      "grad_norm": 16.316425323486328,
+      "learning_rate": 3.9325599999999996e-07,
+      "loss": 0.409,
+      "step": 151750
+    },
+    {
+      "epoch": 1.518,
+      "grad_norm": 3.0078415870666504,
+      "learning_rate": 3.93056e-07,
+      "loss": 0.4389,
+      "step": 151800
+    },
+    {
+      "epoch": 1.5185,
+      "grad_norm": 57.24897384643555,
+      "learning_rate": 3.92856e-07,
+      "loss": 0.4834,
+      "step": 151850
+    },
+    {
+      "epoch": 1.5190000000000001,
+      "grad_norm": 9.619451522827148,
+      "learning_rate": 3.9265599999999997e-07,
+      "loss": 0.4941,
+      "step": 151900
+    },
+    {
+      "epoch": 1.5194999999999999,
+      "grad_norm": 66.24259185791016,
+      "learning_rate": 3.92456e-07,
+      "loss": 0.4241,
+      "step": 151950
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 64.54350280761719,
+      "learning_rate": 3.92256e-07,
+      "loss": 0.4539,
+      "step": 152000
+    },
+    {
+      "epoch": 1.5205,
+      "grad_norm": 3.8642959594726562,
+      "learning_rate": 3.92056e-07,
+      "loss": 0.4306,
+      "step": 152050
+    },
+    {
+      "epoch": 1.521,
+      "grad_norm": 88.06119537353516,
+      "learning_rate": 3.9185599999999996e-07,
+      "loss": 0.4,
+      "step": 152100
+    },
+    {
+      "epoch": 1.5215,
+      "grad_norm": 0.5030847191810608,
+      "learning_rate": 3.91656e-07,
+      "loss": 0.4478,
+      "step": 152150
+    },
+    {
+      "epoch": 1.522,
+      "grad_norm": 74.51904296875,
+      "learning_rate": 3.9145600000000003e-07,
+      "loss": 0.4082,
+      "step": 152200
+    },
+    {
+      "epoch": 1.5225,
+      "grad_norm": 7.973733425140381,
+      "learning_rate": 3.9125599999999997e-07,
+      "loss": 0.4605,
+      "step": 152250
+    },
+    {
+      "epoch": 1.5230000000000001,
+      "grad_norm": 52.102630615234375,
+      "learning_rate": 3.91056e-07,
+      "loss": 0.4287,
+      "step": 152300
+    },
+    {
+      "epoch": 1.5234999999999999,
+      "grad_norm": 10.355042457580566,
+      "learning_rate": 3.90856e-07,
+      "loss": 0.532,
+      "step": 152350
+    },
+    {
+      "epoch": 1.524,
+      "grad_norm": 13.536386489868164,
+      "learning_rate": 3.90656e-07,
+      "loss": 0.3487,
+      "step": 152400
+    },
+    {
+      "epoch": 1.5245,
+      "grad_norm": 17.59664535522461,
+      "learning_rate": 3.90456e-07,
+      "loss": 0.5859,
+      "step": 152450
+    },
+    {
+      "epoch": 1.525,
+      "grad_norm": 37.894744873046875,
+      "learning_rate": 3.90256e-07,
+      "loss": 0.3717,
+      "step": 152500
+    },
+    {
+      "epoch": 1.5255,
+      "grad_norm": 87.29824829101562,
+      "learning_rate": 3.90056e-07,
+      "loss": 0.3377,
+      "step": 152550
+    },
+    {
+      "epoch": 1.526,
+      "grad_norm": 150.49517822265625,
+      "learning_rate": 3.8985599999999997e-07,
+      "loss": 0.4282,
+      "step": 152600
+    },
+    {
+      "epoch": 1.5265,
+      "grad_norm": 22.18712615966797,
+      "learning_rate": 3.89656e-07,
+      "loss": 0.3628,
+      "step": 152650
+    },
+    {
+      "epoch": 1.5270000000000001,
+      "grad_norm": 7.895515441894531,
+      "learning_rate": 3.8945999999999994e-07,
+      "loss": 0.4914,
+      "step": 152700
+    },
+    {
+      "epoch": 1.5274999999999999,
+      "grad_norm": 98.30242919921875,
+      "learning_rate": 3.8926e-07,
+      "loss": 0.4066,
+      "step": 152750
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 71.57797241210938,
+      "learning_rate": 3.8906e-07,
+      "loss": 0.4606,
+      "step": 152800
+    },
+    {
+      "epoch": 1.5285,
+      "grad_norm": 52.828861236572266,
+      "learning_rate": 3.8885999999999995e-07,
+      "loss": 0.4835,
+      "step": 152850
+    },
+    {
+      "epoch": 1.529,
+      "grad_norm": 1.3210022449493408,
+      "learning_rate": 3.8866e-07,
+      "loss": 0.6443,
+      "step": 152900
+    },
+    {
+      "epoch": 1.5295,
+      "grad_norm": 0.20396688580513,
+      "learning_rate": 3.8846e-07,
+      "loss": 0.3461,
+      "step": 152950
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.065512180328369,
+      "learning_rate": 3.8825999999999996e-07,
+      "loss": 0.5639,
+      "step": 153000
+    },
+    {
+      "epoch": 1.5305,
+      "grad_norm": 0.27754420042037964,
+      "learning_rate": 3.8806e-07,
+      "loss": 0.357,
+      "step": 153050
+    },
+    {
+      "epoch": 1.5310000000000001,
+      "grad_norm": 105.62935638427734,
+      "learning_rate": 3.8786e-07,
+      "loss": 0.51,
+      "step": 153100
+    },
+    {
+      "epoch": 1.5314999999999999,
+      "grad_norm": 78.91966247558594,
+      "learning_rate": 3.8765999999999997e-07,
+      "loss": 0.3173,
+      "step": 153150
+    },
+    {
+      "epoch": 1.532,
+      "grad_norm": 22.052091598510742,
+      "learning_rate": 3.8746e-07,
+      "loss": 0.4488,
+      "step": 153200
+    },
+    {
+      "epoch": 1.5325,
+      "grad_norm": 105.67029571533203,
+      "learning_rate": 3.8726e-07,
+      "loss": 0.3544,
+      "step": 153250
+    },
+    {
+      "epoch": 1.533,
+      "grad_norm": 68.83394622802734,
+      "learning_rate": 3.8705999999999997e-07,
+      "loss": 0.4906,
+      "step": 153300
+    },
+    {
+      "epoch": 1.5335,
+      "grad_norm": 35.794700622558594,
+      "learning_rate": 3.8685999999999996e-07,
+      "loss": 0.3602,
+      "step": 153350
+    },
+    {
+      "epoch": 1.534,
+      "grad_norm": 8.180852890014648,
+      "learning_rate": 3.8666e-07,
+      "loss": 0.4035,
+      "step": 153400
+    },
+    {
+      "epoch": 1.5345,
+      "grad_norm": 18.921524047851562,
+      "learning_rate": 3.8646000000000003e-07,
+      "loss": 0.4177,
+      "step": 153450
+    },
+    {
+      "epoch": 1.5350000000000001,
+      "grad_norm": 28.274133682250977,
+      "learning_rate": 3.8625999999999997e-07,
+      "loss": 0.3924,
+      "step": 153500
+    },
+    {
+      "epoch": 1.5354999999999999,
+      "grad_norm": 60.59059524536133,
+      "learning_rate": 3.8606e-07,
+      "loss": 0.641,
+      "step": 153550
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 13.755937576293945,
+      "learning_rate": 3.8586e-07,
+      "loss": 0.3756,
+      "step": 153600
+    },
+    {
+      "epoch": 1.5365,
+      "grad_norm": 28.67469596862793,
+      "learning_rate": 3.8566e-07,
+      "loss": 0.5329,
+      "step": 153650
+    },
+    {
+      "epoch": 1.537,
+      "grad_norm": 67.26752471923828,
+      "learning_rate": 3.8546e-07,
+      "loss": 0.4442,
+      "step": 153700
+    },
+    {
+      "epoch": 1.5375,
+      "grad_norm": 76.13204956054688,
+      "learning_rate": 3.8526e-07,
+      "loss": 0.3637,
+      "step": 153750
+    },
+    {
+      "epoch": 1.538,
+      "grad_norm": 108.6531753540039,
+      "learning_rate": 3.8506e-07,
+      "loss": 0.4456,
+      "step": 153800
+    },
+    {
+      "epoch": 1.5385,
+      "grad_norm": 82.92815399169922,
+      "learning_rate": 3.8485999999999997e-07,
+      "loss": 0.3671,
+      "step": 153850
+    },
+    {
+      "epoch": 1.5390000000000001,
+      "grad_norm": 3.92008376121521,
+      "learning_rate": 3.8466e-07,
+      "loss": 0.354,
+      "step": 153900
+    },
+    {
+      "epoch": 1.5394999999999999,
+      "grad_norm": 46.99858093261719,
+      "learning_rate": 3.8446e-07,
+      "loss": 0.5856,
+      "step": 153950
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.050018310546875,
+      "learning_rate": 3.8426e-07,
+      "loss": 0.5328,
+      "step": 154000
+    },
+    {
+      "epoch": 1.5405,
+      "grad_norm": 59.63269805908203,
+      "learning_rate": 3.8406e-07,
+      "loss": 0.5117,
+      "step": 154050
+    },
+    {
+      "epoch": 1.541,
+      "grad_norm": 12.401506423950195,
+      "learning_rate": 3.8385999999999994e-07,
+      "loss": 0.4722,
+      "step": 154100
+    },
+    {
+      "epoch": 1.5415,
+      "grad_norm": 104.35944366455078,
+      "learning_rate": 3.8366e-07,
+      "loss": 0.5007,
+      "step": 154150
+    },
+    {
+      "epoch": 1.542,
+      "grad_norm": 32.52841567993164,
+      "learning_rate": 3.8346e-07,
+      "loss": 0.4704,
+      "step": 154200
+    },
+    {
+      "epoch": 1.5425,
+      "grad_norm": 101.7193374633789,
+      "learning_rate": 3.8325999999999995e-07,
+      "loss": 0.4382,
+      "step": 154250
+    },
+    {
+      "epoch": 1.5430000000000001,
+      "grad_norm": 89.42278289794922,
+      "learning_rate": 3.8306e-07,
+      "loss": 0.5037,
+      "step": 154300
+    },
+    {
+      "epoch": 1.5434999999999999,
+      "grad_norm": 35.02275466918945,
+      "learning_rate": 3.8286e-07,
+      "loss": 0.4409,
+      "step": 154350
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 78.44393157958984,
+      "learning_rate": 3.8265999999999996e-07,
+      "loss": 0.4195,
+      "step": 154400
+    },
+    {
+      "epoch": 1.5445,
+      "grad_norm": 14.024569511413574,
+      "learning_rate": 3.8246e-07,
+      "loss": 0.3931,
+      "step": 154450
+    },
+    {
+      "epoch": 1.545,
+      "grad_norm": 29.970157623291016,
+      "learning_rate": 3.8226e-07,
+      "loss": 0.4982,
+      "step": 154500
+    },
+    {
+      "epoch": 1.5455,
+      "grad_norm": 92.17186737060547,
+      "learning_rate": 3.8205999999999997e-07,
+      "loss": 0.489,
+      "step": 154550
+    },
+    {
+      "epoch": 1.546,
+      "grad_norm": 103.02266693115234,
+      "learning_rate": 3.8185999999999995e-07,
+      "loss": 0.4268,
+      "step": 154600
+    },
+    {
+      "epoch": 1.5465,
+      "grad_norm": 2.3112690448760986,
+      "learning_rate": 3.8166e-07,
+      "loss": 0.4287,
+      "step": 154650
+    },
+    {
+      "epoch": 1.5470000000000002,
+      "grad_norm": 71.17611694335938,
+      "learning_rate": 3.8146000000000003e-07,
+      "loss": 0.356,
+      "step": 154700
+    },
+    {
+      "epoch": 1.5474999999999999,
+      "grad_norm": 21.637741088867188,
+      "learning_rate": 3.8125999999999996e-07,
+      "loss": 0.4844,
+      "step": 154750
+    },
+    {
+      "epoch": 1.548,
+      "grad_norm": 26.230649948120117,
+      "learning_rate": 3.8106e-07,
+      "loss": 0.5267,
+      "step": 154800
+    },
+    {
+      "epoch": 1.5485,
+      "grad_norm": 91.8410415649414,
+      "learning_rate": 3.8086e-07,
+      "loss": 0.4135,
+      "step": 154850
+    },
+    {
+      "epoch": 1.549,
+      "grad_norm": 12.015192985534668,
+      "learning_rate": 3.8065999999999997e-07,
+      "loss": 0.4938,
+      "step": 154900
+    },
+    {
+      "epoch": 1.5495,
+      "grad_norm": 5.9891533851623535,
+      "learning_rate": 3.8046e-07,
+      "loss": 0.3827,
+      "step": 154950
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.673921525478363,
+      "learning_rate": 3.8026e-07,
+      "loss": 0.4642,
+      "step": 155000
+    },
+    {
+      "epoch": 1.5505,
+      "grad_norm": 48.262916564941406,
+      "learning_rate": 3.8006e-07,
+      "loss": 0.2987,
+      "step": 155050
+    },
+    {
+      "epoch": 1.5510000000000002,
+      "grad_norm": 15.329411506652832,
+      "learning_rate": 3.7985999999999996e-07,
+      "loss": 0.4899,
+      "step": 155100
+    },
+    {
+      "epoch": 1.5514999999999999,
+      "grad_norm": 54.15485763549805,
+      "learning_rate": 3.7966e-07,
+      "loss": 0.3895,
+      "step": 155150
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 76.88223266601562,
+      "learning_rate": 3.7946e-07,
+      "loss": 0.3881,
+      "step": 155200
+    },
+    {
+      "epoch": 1.5525,
+      "grad_norm": 19.991403579711914,
+      "learning_rate": 3.7925999999999997e-07,
+      "loss": 0.398,
+      "step": 155250
+    },
+    {
+      "epoch": 1.553,
+      "grad_norm": 94.80240631103516,
+      "learning_rate": 3.7906e-07,
+      "loss": 0.4081,
+      "step": 155300
+    },
+    {
+      "epoch": 1.5535,
+      "grad_norm": 122.23736572265625,
+      "learning_rate": 3.7885999999999994e-07,
+      "loss": 0.3894,
+      "step": 155350
+    },
+    {
+      "epoch": 1.554,
+      "grad_norm": 57.91024398803711,
+      "learning_rate": 3.7866e-07,
+      "loss": 0.4474,
+      "step": 155400
+    },
+    {
+      "epoch": 1.5545,
+      "grad_norm": 8.4732666015625,
+      "learning_rate": 3.7846e-07,
+      "loss": 0.4155,
+      "step": 155450
+    },
+    {
+      "epoch": 1.5550000000000002,
+      "grad_norm": 12.340983390808105,
+      "learning_rate": 3.7825999999999995e-07,
+      "loss": 0.4288,
+      "step": 155500
+    },
+    {
+      "epoch": 1.5554999999999999,
+      "grad_norm": 13.673118591308594,
+      "learning_rate": 3.7806e-07,
+      "loss": 0.5154,
+      "step": 155550
+    },
+    {
+      "epoch": 1.556,
+      "grad_norm": 112.43163299560547,
+      "learning_rate": 3.7785999999999997e-07,
+      "loss": 0.3473,
+      "step": 155600
+    },
+    {
+      "epoch": 1.5565,
+      "grad_norm": 10.982392311096191,
+      "learning_rate": 3.7765999999999996e-07,
+      "loss": 0.3726,
+      "step": 155650
+    },
+    {
+      "epoch": 1.557,
+      "grad_norm": 105.76234436035156,
+      "learning_rate": 3.7746e-07,
+      "loss": 0.3479,
+      "step": 155700
+    },
+    {
+      "epoch": 1.5575,
+      "grad_norm": 76.24043273925781,
+      "learning_rate": 3.7726e-07,
+      "loss": 0.4958,
+      "step": 155750
+    },
+    {
+      "epoch": 1.558,
+      "grad_norm": 6.342814922332764,
+      "learning_rate": 3.7705999999999996e-07,
+      "loss": 0.3893,
+      "step": 155800
+    },
+    {
+      "epoch": 1.5585,
+      "grad_norm": 74.10182189941406,
+      "learning_rate": 3.7685999999999995e-07,
+      "loss": 0.4655,
+      "step": 155850
+    },
+    {
+      "epoch": 1.5590000000000002,
+      "grad_norm": 0.016593320295214653,
+      "learning_rate": 3.7666e-07,
+      "loss": 0.4724,
+      "step": 155900
+    },
+    {
+      "epoch": 1.5594999999999999,
+      "grad_norm": 3.463813304901123,
+      "learning_rate": 3.7646e-07,
+      "loss": 0.3882,
+      "step": 155950
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 28.10302734375,
+      "learning_rate": 3.7625999999999996e-07,
+      "loss": 0.4819,
+      "step": 156000
+    },
+    {
+      "epoch": 1.5605,
+      "grad_norm": 90.9014892578125,
+      "learning_rate": 3.7606e-07,
+      "loss": 0.5103,
+      "step": 156050
+    },
+    {
+      "epoch": 1.561,
+      "grad_norm": 5.384343147277832,
+      "learning_rate": 3.7586000000000003e-07,
+      "loss": 0.3814,
+      "step": 156100
+    },
+    {
+      "epoch": 1.5615,
+      "grad_norm": 63.378196716308594,
+      "learning_rate": 3.7565999999999996e-07,
+      "loss": 0.3121,
+      "step": 156150
+    },
+    {
+      "epoch": 1.562,
+      "grad_norm": 67.78466796875,
+      "learning_rate": 3.7546e-07,
+      "loss": 0.3317,
+      "step": 156200
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 38.204132080078125,
+      "learning_rate": 3.7526e-07,
+      "loss": 0.6073,
+      "step": 156250
+    },
+    {
+      "epoch": 1.563,
+      "grad_norm": 29.50925636291504,
+      "learning_rate": 3.7505999999999997e-07,
+      "loss": 0.3818,
+      "step": 156300
+    },
+    {
+      "epoch": 1.5635,
+      "grad_norm": 55.05265426635742,
+      "learning_rate": 3.7486e-07,
+      "loss": 0.4596,
+      "step": 156350
+    },
+    {
+      "epoch": 1.564,
+      "grad_norm": 124.0997085571289,
+      "learning_rate": 3.7466e-07,
+      "loss": 0.4736,
+      "step": 156400
+    },
+    {
+      "epoch": 1.5645,
+      "grad_norm": 38.588619232177734,
+      "learning_rate": 3.7446e-07,
+      "loss": 0.4511,
+      "step": 156450
+    },
+    {
+      "epoch": 1.565,
+      "grad_norm": 80.09603118896484,
+      "learning_rate": 3.7425999999999996e-07,
+      "loss": 0.4403,
+      "step": 156500
+    },
+    {
+      "epoch": 1.5655000000000001,
+      "grad_norm": 1.9069684743881226,
+      "learning_rate": 3.7406e-07,
+      "loss": 0.5459,
+      "step": 156550
+    },
+    {
+      "epoch": 1.5659999999999998,
+      "grad_norm": 1.7578861713409424,
+      "learning_rate": 3.7386e-07,
+      "loss": 0.5356,
+      "step": 156600
+    },
+    {
+      "epoch": 1.5665,
+      "grad_norm": 13.54135513305664,
+      "learning_rate": 3.7365999999999997e-07,
+      "loss": 0.4722,
+      "step": 156650
+    },
+    {
+      "epoch": 1.567,
+      "grad_norm": 42.47798156738281,
+      "learning_rate": 3.7346e-07,
+      "loss": 0.4462,
+      "step": 156700
+    },
+    {
+      "epoch": 1.5675,
+      "grad_norm": 32.1161003112793,
+      "learning_rate": 3.7326399999999995e-07,
+      "loss": 0.4484,
+      "step": 156750
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 44.59100341796875,
+      "learning_rate": 3.73064e-07,
+      "loss": 0.3855,
+      "step": 156800
+    },
+    {
+      "epoch": 1.5685,
+      "grad_norm": 15.064719200134277,
+      "learning_rate": 3.7286399999999997e-07,
+      "loss": 0.4979,
+      "step": 156850
+    },
+    {
+      "epoch": 1.569,
+      "grad_norm": 82.406982421875,
+      "learning_rate": 3.7266399999999995e-07,
+      "loss": 0.3934,
+      "step": 156900
+    },
+    {
+      "epoch": 1.5695000000000001,
+      "grad_norm": 1.5106115341186523,
+      "learning_rate": 3.72464e-07,
+      "loss": 0.4888,
+      "step": 156950
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 34.84396743774414,
+      "learning_rate": 3.72264e-07,
+      "loss": 0.3313,
+      "step": 157000
+    },
+    {
+      "epoch": 1.5705,
+      "grad_norm": 56.88155746459961,
+      "learning_rate": 3.7206399999999996e-07,
+      "loss": 0.6379,
+      "step": 157050
+    },
+    {
+      "epoch": 1.571,
+      "grad_norm": 47.09871292114258,
+      "learning_rate": 3.7186399999999995e-07,
+      "loss": 0.4519,
+      "step": 157100
+    },
+    {
+      "epoch": 1.5715,
+      "grad_norm": 44.833335876464844,
+      "learning_rate": 3.71664e-07,
+      "loss": 0.3936,
+      "step": 157150
+    },
+    {
+      "epoch": 1.572,
+      "grad_norm": 85.64391326904297,
+      "learning_rate": 3.71464e-07,
+      "loss": 0.3674,
+      "step": 157200
+    },
+    {
+      "epoch": 1.5725,
+      "grad_norm": 69.18328857421875,
+      "learning_rate": 3.7126399999999996e-07,
+      "loss": 0.5126,
+      "step": 157250
+    },
+    {
+      "epoch": 1.573,
+      "grad_norm": 98.84770202636719,
+      "learning_rate": 3.71064e-07,
+      "loss": 0.4173,
+      "step": 157300
+    },
+    {
+      "epoch": 1.5735000000000001,
+      "grad_norm": 22.579782485961914,
+      "learning_rate": 3.7086400000000003e-07,
+      "loss": 0.4732,
+      "step": 157350
+    },
+    {
+      "epoch": 1.5739999999999998,
+      "grad_norm": 27.351652145385742,
+      "learning_rate": 3.7066399999999996e-07,
+      "loss": 0.5637,
+      "step": 157400
+    },
+    {
+      "epoch": 1.5745,
+      "grad_norm": 66.01678466796875,
+      "learning_rate": 3.70464e-07,
+      "loss": 0.3713,
+      "step": 157450
+    },
+    {
+      "epoch": 1.575,
+      "grad_norm": 4.506946563720703,
+      "learning_rate": 3.70264e-07,
+      "loss": 0.2717,
+      "step": 157500
+    },
+    {
+      "epoch": 1.5755,
+      "grad_norm": 61.51058578491211,
+      "learning_rate": 3.7006399999999997e-07,
+      "loss": 0.4417,
+      "step": 157550
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 3.467181921005249,
+      "learning_rate": 3.69864e-07,
+      "loss": 0.4313,
+      "step": 157600
+    },
+    {
+      "epoch": 1.5765,
+      "grad_norm": 14.573429107666016,
+      "learning_rate": 3.69664e-07,
+      "loss": 0.4639,
+      "step": 157650
+    },
+    {
+      "epoch": 1.577,
+      "grad_norm": 54.943565368652344,
+      "learning_rate": 3.69464e-07,
+      "loss": 0.428,
+      "step": 157700
+    },
+    {
+      "epoch": 1.5775000000000001,
+      "grad_norm": 8.155677795410156,
+      "learning_rate": 3.6926799999999997e-07,
+      "loss": 0.4534,
+      "step": 157750
+    },
+    {
+      "epoch": 1.5779999999999998,
+      "grad_norm": 67.54132080078125,
+      "learning_rate": 3.69068e-07,
+      "loss": 0.4839,
+      "step": 157800
+    },
+    {
+      "epoch": 1.5785,
+      "grad_norm": 10.213589668273926,
+      "learning_rate": 3.6886799999999994e-07,
+      "loss": 0.3556,
+      "step": 157850
+    },
+    {
+      "epoch": 1.579,
+      "grad_norm": 109.822021484375,
+      "learning_rate": 3.68668e-07,
+      "loss": 0.4593,
+      "step": 157900
+    },
+    {
+      "epoch": 1.5795,
+      "grad_norm": 82.01921081542969,
+      "learning_rate": 3.68468e-07,
+      "loss": 0.3996,
+      "step": 157950
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 60.44963836669922,
+      "learning_rate": 3.6826799999999995e-07,
+      "loss": 0.5142,
+      "step": 158000
+    },
+    {
+      "epoch": 1.5805,
+      "grad_norm": 101.18560791015625,
+      "learning_rate": 3.68068e-07,
+      "loss": 0.39,
+      "step": 158050
+    },
+    {
+      "epoch": 1.581,
+      "grad_norm": 73.67401123046875,
+      "learning_rate": 3.6786799999999997e-07,
+      "loss": 0.4294,
+      "step": 158100
+    },
+    {
+      "epoch": 1.5815000000000001,
+      "grad_norm": 38.87802505493164,
+      "learning_rate": 3.6766799999999995e-07,
+      "loss": 0.4877,
+      "step": 158150
+    },
+    {
+      "epoch": 1.5819999999999999,
+      "grad_norm": 36.239723205566406,
+      "learning_rate": 3.67468e-07,
+      "loss": 0.3294,
+      "step": 158200
+    },
+    {
+      "epoch": 1.5825,
+      "grad_norm": 16.11030387878418,
+      "learning_rate": 3.67268e-07,
+      "loss": 0.5534,
+      "step": 158250
+    },
+    {
+      "epoch": 1.583,
+      "grad_norm": 43.37452697753906,
+      "learning_rate": 3.67068e-07,
+      "loss": 0.4388,
+      "step": 158300
+    },
+    {
+      "epoch": 1.5835,
+      "grad_norm": 15.790046691894531,
+      "learning_rate": 3.66868e-07,
+      "loss": 0.4595,
+      "step": 158350
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 49.98896026611328,
+      "learning_rate": 3.66668e-07,
+      "loss": 0.598,
+      "step": 158400
+    },
+    {
+      "epoch": 1.5845,
+      "grad_norm": 151.6587371826172,
+      "learning_rate": 3.66468e-07,
+      "loss": 0.5676,
+      "step": 158450
+    },
+    {
+      "epoch": 1.585,
+      "grad_norm": 150.25958251953125,
+      "learning_rate": 3.6626799999999995e-07,
+      "loss": 0.4359,
+      "step": 158500
+    },
+    {
+      "epoch": 1.5855000000000001,
+      "grad_norm": 79.11650085449219,
+      "learning_rate": 3.66068e-07,
+      "loss": 0.4847,
+      "step": 158550
+    },
+    {
+      "epoch": 1.5859999999999999,
+      "grad_norm": 16.014951705932617,
+      "learning_rate": 3.6586800000000003e-07,
+      "loss": 0.4409,
+      "step": 158600
+    },
+    {
+      "epoch": 1.5865,
+      "grad_norm": 38.4268798828125,
+      "learning_rate": 3.6566799999999996e-07,
+      "loss": 0.4406,
+      "step": 158650
+    },
+    {
+      "epoch": 1.587,
+      "grad_norm": 75.6319808959961,
+      "learning_rate": 3.65468e-07,
+      "loss": 0.4073,
+      "step": 158700
+    },
+    {
+      "epoch": 1.5875,
+      "grad_norm": 83.09989166259766,
+      "learning_rate": 3.65268e-07,
+      "loss": 0.6108,
+      "step": 158750
+    },
+    {
+      "epoch": 1.588,
+      "grad_norm": 85.7934341430664,
+      "learning_rate": 3.6506799999999997e-07,
+      "loss": 0.4062,
+      "step": 158800
+    },
+    {
+      "epoch": 1.5885,
+      "grad_norm": 63.43593978881836,
+      "learning_rate": 3.64868e-07,
+      "loss": 0.4971,
+      "step": 158850
+    },
+    {
+      "epoch": 1.589,
+      "grad_norm": 0.3882288336753845,
+      "learning_rate": 3.64668e-07,
+      "loss": 0.4136,
+      "step": 158900
+    },
+    {
+      "epoch": 1.5895000000000001,
+      "grad_norm": 69.4052963256836,
+      "learning_rate": 3.64468e-07,
+      "loss": 0.5876,
+      "step": 158950
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 0.12768994271755219,
+      "learning_rate": 3.6426799999999996e-07,
+      "loss": 0.5236,
+      "step": 159000
+    },
+    {
+      "epoch": 1.5905,
+      "grad_norm": 123.2761459350586,
+      "learning_rate": 3.64068e-07,
+      "loss": 0.503,
+      "step": 159050
+    },
+    {
+      "epoch": 1.591,
+      "grad_norm": 0.08437743782997131,
+      "learning_rate": 3.63868e-07,
+      "loss": 0.5431,
+      "step": 159100
+    },
+    {
+      "epoch": 1.5915,
+      "grad_norm": 50.3804817199707,
+      "learning_rate": 3.6366799999999997e-07,
+      "loss": 0.3803,
+      "step": 159150
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 66.29913330078125,
+      "learning_rate": 3.63468e-07,
+      "loss": 0.441,
+      "step": 159200
+    },
+    {
+      "epoch": 1.5925,
+      "grad_norm": 57.80009460449219,
+      "learning_rate": 3.6326799999999994e-07,
+      "loss": 0.5003,
+      "step": 159250
+    },
+    {
+      "epoch": 1.593,
+      "grad_norm": 38.19333267211914,
+      "learning_rate": 3.63068e-07,
+      "loss": 0.3942,
+      "step": 159300
+    },
+    {
+      "epoch": 1.5935000000000001,
+      "grad_norm": 108.0379638671875,
+      "learning_rate": 3.62868e-07,
+      "loss": 0.5143,
+      "step": 159350
+    },
+    {
+      "epoch": 1.5939999999999999,
+      "grad_norm": 37.051204681396484,
+      "learning_rate": 3.6266799999999995e-07,
+      "loss": 0.4615,
+      "step": 159400
+    },
+    {
+      "epoch": 1.5945,
+      "grad_norm": 32.642669677734375,
+      "learning_rate": 3.62468e-07,
+      "loss": 0.4396,
+      "step": 159450
+    },
+    {
+      "epoch": 1.595,
+      "grad_norm": 64.99166107177734,
+      "learning_rate": 3.6226799999999997e-07,
+      "loss": 0.4363,
+      "step": 159500
+    },
+    {
+      "epoch": 1.5955,
+      "grad_norm": 57.88900375366211,
+      "learning_rate": 3.62068e-07,
+      "loss": 0.4541,
+      "step": 159550
+    },
+    {
+      "epoch": 1.596,
+      "grad_norm": 0.2389134019613266,
+      "learning_rate": 3.61868e-07,
+      "loss": 0.5371,
+      "step": 159600
+    },
+    {
+      "epoch": 1.5965,
+      "grad_norm": 7.437679290771484,
+      "learning_rate": 3.61668e-07,
+      "loss": 0.4728,
+      "step": 159650
+    },
+    {
+      "epoch": 1.597,
+      "grad_norm": 78.19436645507812,
+      "learning_rate": 3.61468e-07,
+      "loss": 0.4397,
+      "step": 159700
+    },
+    {
+      "epoch": 1.5975000000000001,
+      "grad_norm": 110.06568908691406,
+      "learning_rate": 3.6127199999999995e-07,
+      "loss": 0.4691,
+      "step": 159750
+    },
+    {
+      "epoch": 1.5979999999999999,
+      "grad_norm": 67.42326354980469,
+      "learning_rate": 3.61072e-07,
+      "loss": 0.3647,
+      "step": 159800
+    },
+    {
+      "epoch": 1.5985,
+      "grad_norm": 82.26587677001953,
+      "learning_rate": 3.6087200000000003e-07,
+      "loss": 0.5464,
+      "step": 159850
+    },
+    {
+      "epoch": 1.599,
+      "grad_norm": 92.55001068115234,
+      "learning_rate": 3.6067199999999996e-07,
+      "loss": 0.3611,
+      "step": 159900
+    },
+    {
+      "epoch": 1.5995,
+      "grad_norm": 35.45196533203125,
+      "learning_rate": 3.60472e-07,
+      "loss": 0.3369,
+      "step": 159950
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.73994445800781,
+      "learning_rate": 3.60272e-07,
+      "loss": 0.3748,
+      "step": 160000
+    },
+    {
+      "epoch": 1.6005,
+      "grad_norm": 2.0758419036865234,
+      "learning_rate": 3.6007199999999997e-07,
+      "loss": 0.4442,
+      "step": 160050
+    },
+    {
+      "epoch": 1.601,
+      "grad_norm": 62.947288513183594,
+      "learning_rate": 3.59872e-07,
+      "loss": 0.4279,
+      "step": 160100
+    },
+    {
+      "epoch": 1.6015000000000001,
+      "grad_norm": 0.5712276101112366,
+      "learning_rate": 3.59672e-07,
+      "loss": 0.4415,
+      "step": 160150
+    },
+    {
+      "epoch": 1.6019999999999999,
+      "grad_norm": 31.01521873474121,
+      "learning_rate": 3.59472e-07,
+      "loss": 0.4596,
+      "step": 160200
+    },
+    {
+      "epoch": 1.6025,
+      "grad_norm": 58.51720428466797,
+      "learning_rate": 3.5927199999999996e-07,
+      "loss": 0.4667,
+      "step": 160250
+    },
+    {
+      "epoch": 1.603,
+      "grad_norm": 23.069557189941406,
+      "learning_rate": 3.59072e-07,
+      "loss": 0.3335,
+      "step": 160300
+    },
+    {
+      "epoch": 1.6035,
+      "grad_norm": 93.3023910522461,
+      "learning_rate": 3.58872e-07,
+      "loss": 0.519,
+      "step": 160350
+    },
+    {
+      "epoch": 1.604,
+      "grad_norm": 5.520042896270752,
+      "learning_rate": 3.5867199999999997e-07,
+      "loss": 0.42,
+      "step": 160400
+    },
+    {
+      "epoch": 1.6045,
+      "grad_norm": 49.47138214111328,
+      "learning_rate": 3.58472e-07,
+      "loss": 0.5085,
+      "step": 160450
+    },
+    {
+      "epoch": 1.605,
+      "grad_norm": 93.62559509277344,
+      "learning_rate": 3.5827199999999994e-07,
+      "loss": 0.3626,
+      "step": 160500
+    },
+    {
+      "epoch": 1.6055000000000001,
+      "grad_norm": 28.985837936401367,
+      "learning_rate": 3.58072e-07,
+      "loss": 0.4725,
+      "step": 160550
+    },
+    {
+      "epoch": 1.6059999999999999,
+      "grad_norm": 0.5782918334007263,
+      "learning_rate": 3.57872e-07,
+      "loss": 0.344,
+      "step": 160600
+    },
+    {
+      "epoch": 1.6065,
+      "grad_norm": 62.849220275878906,
+      "learning_rate": 3.5767199999999995e-07,
+      "loss": 0.3763,
+      "step": 160650
+    },
+    {
+      "epoch": 1.607,
+      "grad_norm": 62.66322708129883,
+      "learning_rate": 3.57472e-07,
+      "loss": 0.4729,
+      "step": 160700
+    },
+    {
+      "epoch": 1.6075,
+      "grad_norm": 1.8875168561935425,
+      "learning_rate": 3.5727199999999997e-07,
+      "loss": 0.4565,
+      "step": 160750
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 86.59226989746094,
+      "learning_rate": 3.57072e-07,
+      "loss": 0.3441,
+      "step": 160800
+    },
+    {
+      "epoch": 1.6085,
+      "grad_norm": 22.594993591308594,
+      "learning_rate": 3.56872e-07,
+      "loss": 0.5117,
+      "step": 160850
+    },
+    {
+      "epoch": 1.609,
+      "grad_norm": 31.01827049255371,
+      "learning_rate": 3.56672e-07,
+      "loss": 0.4471,
+      "step": 160900
+    },
+    {
+      "epoch": 1.6095000000000002,
+      "grad_norm": 85.61234283447266,
+      "learning_rate": 3.56472e-07,
+      "loss": 0.3757,
+      "step": 160950
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 87.01544189453125,
+      "learning_rate": 3.5627199999999995e-07,
+      "loss": 0.5461,
+      "step": 161000
+    },
+    {
+      "epoch": 1.6105,
+      "grad_norm": 4.279980659484863,
+      "learning_rate": 3.56072e-07,
+      "loss": 0.4856,
+      "step": 161050
+    },
+    {
+      "epoch": 1.611,
+      "grad_norm": 121.6936264038086,
+      "learning_rate": 3.55872e-07,
+      "loss": 0.5672,
+      "step": 161100
+    },
+    {
+      "epoch": 1.6115,
+      "grad_norm": 12.646116256713867,
+      "learning_rate": 3.5567199999999996e-07,
+      "loss": 0.4158,
+      "step": 161150
+    },
+    {
+      "epoch": 1.612,
+      "grad_norm": 166.2145233154297,
+      "learning_rate": 3.55472e-07,
+      "loss": 0.4199,
+      "step": 161200
+    },
+    {
+      "epoch": 1.6125,
+      "grad_norm": 37.84153366088867,
+      "learning_rate": 3.55272e-07,
+      "loss": 0.5097,
+      "step": 161250
+    },
+    {
+      "epoch": 1.613,
+      "grad_norm": 21.25974464416504,
+      "learning_rate": 3.5507199999999996e-07,
+      "loss": 0.4476,
+      "step": 161300
+    },
+    {
+      "epoch": 1.6135000000000002,
+      "grad_norm": 1.0741132497787476,
+      "learning_rate": 3.54872e-07,
+      "loss": 0.3204,
+      "step": 161350
+    },
+    {
+      "epoch": 1.6139999999999999,
+      "grad_norm": 76.07339477539062,
+      "learning_rate": 3.54672e-07,
+      "loss": 0.4677,
+      "step": 161400
+    },
+    {
+      "epoch": 1.6145,
+      "grad_norm": 37.11135482788086,
+      "learning_rate": 3.5447199999999997e-07,
+      "loss": 0.4673,
+      "step": 161450
+    },
+    {
+      "epoch": 1.615,
+      "grad_norm": 29.35950469970703,
+      "learning_rate": 3.5427199999999996e-07,
+      "loss": 0.4879,
+      "step": 161500
+    },
+    {
+      "epoch": 1.6155,
+      "grad_norm": 6.429986000061035,
+      "learning_rate": 3.54072e-07,
+      "loss": 0.4428,
+      "step": 161550
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 4.6186933517456055,
+      "learning_rate": 3.53872e-07,
+      "loss": 0.3503,
+      "step": 161600
+    },
+    {
+      "epoch": 1.6165,
+      "grad_norm": 102.59295654296875,
+      "learning_rate": 3.5367199999999997e-07,
+      "loss": 0.3102,
+      "step": 161650
+    },
+    {
+      "epoch": 1.617,
+      "grad_norm": 0.6536028981208801,
+      "learning_rate": 3.53472e-07,
+      "loss": 0.4446,
+      "step": 161700
+    },
+    {
+      "epoch": 1.6175000000000002,
+      "grad_norm": 25.664196014404297,
+      "learning_rate": 3.5327199999999994e-07,
+      "loss": 0.3685,
+      "step": 161750
+    },
+    {
+      "epoch": 1.6179999999999999,
+      "grad_norm": 59.3905143737793,
+      "learning_rate": 3.53072e-07,
+      "loss": 0.5044,
+      "step": 161800
+    },
+    {
+      "epoch": 1.6185,
+      "grad_norm": 35.49124526977539,
+      "learning_rate": 3.52872e-07,
+      "loss": 0.3558,
+      "step": 161850
+    },
+    {
+      "epoch": 1.619,
+      "grad_norm": 36.11788558959961,
+      "learning_rate": 3.5267199999999994e-07,
+      "loss": 0.3282,
+      "step": 161900
+    },
+    {
+      "epoch": 1.6195,
+      "grad_norm": 36.98171615600586,
+      "learning_rate": 3.52472e-07,
+      "loss": 0.4198,
+      "step": 161950
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.539791584014893,
+      "learning_rate": 3.5227199999999997e-07,
+      "loss": 0.4764,
+      "step": 162000
+    },
+    {
+      "epoch": 1.6205,
+      "grad_norm": 5.743781566619873,
+      "learning_rate": 3.52072e-07,
+      "loss": 0.3213,
+      "step": 162050
+    },
+    {
+      "epoch": 1.621,
+      "grad_norm": 1.3049453496932983,
+      "learning_rate": 3.51872e-07,
+      "loss": 0.2321,
+      "step": 162100
+    },
+    {
+      "epoch": 1.6215000000000002,
+      "grad_norm": 17.184179306030273,
+      "learning_rate": 3.51676e-07,
+      "loss": 0.4108,
+      "step": 162150
+    },
+    {
+      "epoch": 1.6219999999999999,
+      "grad_norm": 8.404489517211914,
+      "learning_rate": 3.51476e-07,
+      "loss": 0.4013,
+      "step": 162200
+    },
+    {
+      "epoch": 1.6225,
+      "grad_norm": 4.2238993644714355,
+      "learning_rate": 3.5127599999999995e-07,
+      "loss": 0.413,
+      "step": 162250
+    },
+    {
+      "epoch": 1.623,
+      "grad_norm": 53.541168212890625,
+      "learning_rate": 3.51076e-07,
+      "loss": 0.4472,
+      "step": 162300
+    },
+    {
+      "epoch": 1.6235,
+      "grad_norm": 40.160789489746094,
+      "learning_rate": 3.50876e-07,
+      "loss": 0.3242,
+      "step": 162350
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 2.3333935737609863,
+      "learning_rate": 3.5067599999999996e-07,
+      "loss": 0.4453,
+      "step": 162400
+    },
+    {
+      "epoch": 1.6245,
+      "grad_norm": 28.548282623291016,
+      "learning_rate": 3.50476e-07,
+      "loss": 0.4291,
+      "step": 162450
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 3.4190590381622314,
+      "learning_rate": 3.50276e-07,
+      "loss": 0.3052,
+      "step": 162500
+    },
+    {
+      "epoch": 1.6255,
+      "grad_norm": 50.14396286010742,
+      "learning_rate": 3.5007599999999996e-07,
+      "loss": 0.4178,
+      "step": 162550
+    },
+    {
+      "epoch": 1.626,
+      "grad_norm": 82.80094146728516,
+      "learning_rate": 3.49876e-07,
+      "loss": 0.4882,
+      "step": 162600
+    },
+    {
+      "epoch": 1.6265,
+      "grad_norm": 274.3426208496094,
+      "learning_rate": 3.49676e-07,
+      "loss": 0.4842,
+      "step": 162650
+    },
+    {
+      "epoch": 1.627,
+      "grad_norm": 40.102027893066406,
+      "learning_rate": 3.4947599999999997e-07,
+      "loss": 0.4315,
+      "step": 162700
+    },
+    {
+      "epoch": 1.6275,
+      "grad_norm": 10.764528274536133,
+      "learning_rate": 3.4927599999999996e-07,
+      "loss": 0.3907,
+      "step": 162750
+    },
+    {
+      "epoch": 1.6280000000000001,
+      "grad_norm": 58.64040756225586,
+      "learning_rate": 3.4908e-07,
+      "loss": 0.5452,
+      "step": 162800
+    },
+    {
+      "epoch": 1.6284999999999998,
+      "grad_norm": 80.29693603515625,
+      "learning_rate": 3.4888e-07,
+      "loss": 0.4522,
+      "step": 162850
+    },
+    {
+      "epoch": 1.629,
+      "grad_norm": 79.97079467773438,
+      "learning_rate": 3.4867999999999997e-07,
+      "loss": 0.5351,
+      "step": 162900
+    },
+    {
+      "epoch": 1.6295,
+      "grad_norm": 104.74796295166016,
+      "learning_rate": 3.4848e-07,
+      "loss": 0.5425,
+      "step": 162950
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.748071193695068,
+      "learning_rate": 3.4827999999999994e-07,
+      "loss": 0.3334,
+      "step": 163000
+    },
+    {
+      "epoch": 1.6305,
+      "grad_norm": 167.40061950683594,
+      "learning_rate": 3.4808e-07,
+      "loss": 0.4874,
+      "step": 163050
+    },
+    {
+      "epoch": 1.631,
+      "grad_norm": 0.46852225065231323,
+      "learning_rate": 3.4788e-07,
+      "loss": 0.4224,
+      "step": 163100
+    },
+    {
+      "epoch": 1.6315,
+      "grad_norm": 86.6996078491211,
+      "learning_rate": 3.4767999999999995e-07,
+      "loss": 0.572,
+      "step": 163150
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 17.278841018676758,
+      "learning_rate": 3.4748e-07,
+      "loss": 0.521,
+      "step": 163200
+    },
+    {
+      "epoch": 1.6324999999999998,
+      "grad_norm": 26.78488540649414,
+      "learning_rate": 3.4727999999999997e-07,
+      "loss": 0.4149,
+      "step": 163250
+    },
+    {
+      "epoch": 1.633,
+      "grad_norm": 13.343032836914062,
+      "learning_rate": 3.4708e-07,
+      "loss": 0.3371,
+      "step": 163300
+    },
+    {
+      "epoch": 1.6335,
+      "grad_norm": 0.05451102554798126,
+      "learning_rate": 3.4688e-07,
+      "loss": 0.4209,
+      "step": 163350
+    },
+    {
+      "epoch": 1.634,
+      "grad_norm": 67.80818939208984,
+      "learning_rate": 3.4668e-07,
+      "loss": 0.3541,
+      "step": 163400
+    },
+    {
+      "epoch": 1.6345,
+      "grad_norm": 77.798828125,
+      "learning_rate": 3.4648e-07,
+      "loss": 0.4772,
+      "step": 163450
+    },
+    {
+      "epoch": 1.635,
+      "grad_norm": 70.70539855957031,
+      "learning_rate": 3.4627999999999995e-07,
+      "loss": 0.4017,
+      "step": 163500
+    },
+    {
+      "epoch": 1.6355,
+      "grad_norm": 89.26293182373047,
+      "learning_rate": 3.4608e-07,
+      "loss": 0.5131,
+      "step": 163550
+    },
+    {
+      "epoch": 1.6360000000000001,
+      "grad_norm": 68.28092956542969,
+      "learning_rate": 3.4588e-07,
+      "loss": 0.4411,
+      "step": 163600
+    },
+    {
+      "epoch": 1.6364999999999998,
+      "grad_norm": 14.089746475219727,
+      "learning_rate": 3.4567999999999996e-07,
+      "loss": 0.5162,
+      "step": 163650
+    },
+    {
+      "epoch": 1.637,
+      "grad_norm": 87.18695068359375,
+      "learning_rate": 3.4548e-07,
+      "loss": 0.4253,
+      "step": 163700
+    },
+    {
+      "epoch": 1.6375,
+      "grad_norm": 7.904430389404297,
+      "learning_rate": 3.4528e-07,
+      "loss": 0.4666,
+      "step": 163750
+    },
+    {
+      "epoch": 1.638,
+      "grad_norm": 1.7455227375030518,
+      "learning_rate": 3.4507999999999996e-07,
+      "loss": 0.286,
+      "step": 163800
+    },
+    {
+      "epoch": 1.6385,
+      "grad_norm": 26.539167404174805,
+      "learning_rate": 3.4488e-07,
+      "loss": 0.5327,
+      "step": 163850
+    },
+    {
+      "epoch": 1.639,
+      "grad_norm": 10.772929191589355,
+      "learning_rate": 3.4468e-07,
+      "loss": 0.3932,
+      "step": 163900
+    },
+    {
+      "epoch": 1.6395,
+      "grad_norm": 37.356571197509766,
+      "learning_rate": 3.4447999999999997e-07,
+      "loss": 0.5795,
+      "step": 163950
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 37.573543548583984,
+      "learning_rate": 3.4427999999999996e-07,
+      "loss": 0.5033,
+      "step": 164000
+    },
+    {
+      "epoch": 1.6404999999999998,
+      "grad_norm": 4.547316074371338,
+      "learning_rate": 3.4408e-07,
+      "loss": 0.4576,
+      "step": 164050
+    },
+    {
+      "epoch": 1.641,
+      "grad_norm": 70.57447814941406,
+      "learning_rate": 3.4388e-07,
+      "loss": 0.4713,
+      "step": 164100
+    },
+    {
+      "epoch": 1.6415,
+      "grad_norm": 5.259047508239746,
+      "learning_rate": 3.4367999999999996e-07,
+      "loss": 0.5499,
+      "step": 164150
+    },
+    {
+      "epoch": 1.642,
+      "grad_norm": 10.574108123779297,
+      "learning_rate": 3.4348e-07,
+      "loss": 0.2849,
+      "step": 164200
+    },
+    {
+      "epoch": 1.6425,
+      "grad_norm": 69.39708709716797,
+      "learning_rate": 3.4327999999999993e-07,
+      "loss": 0.4958,
+      "step": 164250
+    },
+    {
+      "epoch": 1.643,
+      "grad_norm": 21.05078887939453,
+      "learning_rate": 3.4307999999999997e-07,
+      "loss": 0.6435,
+      "step": 164300
+    },
+    {
+      "epoch": 1.6435,
+      "grad_norm": 41.12791442871094,
+      "learning_rate": 3.4288e-07,
+      "loss": 0.3685,
+      "step": 164350
+    },
+    {
+      "epoch": 1.6440000000000001,
+      "grad_norm": 118.48395538330078,
+      "learning_rate": 3.4268e-07,
+      "loss": 0.4217,
+      "step": 164400
+    },
+    {
+      "epoch": 1.6444999999999999,
+      "grad_norm": 5.526557922363281,
+      "learning_rate": 3.4248e-07,
+      "loss": 0.3102,
+      "step": 164450
+    },
+    {
+      "epoch": 1.645,
+      "grad_norm": 52.724700927734375,
+      "learning_rate": 3.4227999999999997e-07,
+      "loss": 0.4322,
+      "step": 164500
+    },
+    {
+      "epoch": 1.6455,
+      "grad_norm": 5.8467631340026855,
+      "learning_rate": 3.4208e-07,
+      "loss": 0.4841,
+      "step": 164550
+    },
+    {
+      "epoch": 1.646,
+      "grad_norm": 8.807022094726562,
+      "learning_rate": 3.4188e-07,
+      "loss": 0.3826,
+      "step": 164600
+    },
+    {
+      "epoch": 1.6465,
+      "grad_norm": 0.2374558299779892,
+      "learning_rate": 3.4167999999999997e-07,
+      "loss": 0.4602,
+      "step": 164650
+    },
+    {
+      "epoch": 1.647,
+      "grad_norm": 4.015384197235107,
+      "learning_rate": 3.4148e-07,
+      "loss": 0.4809,
+      "step": 164700
+    },
+    {
+      "epoch": 1.6475,
+      "grad_norm": 0.15211087465286255,
+      "learning_rate": 3.4127999999999994e-07,
+      "loss": 0.3981,
+      "step": 164750
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 62.6292839050293,
+      "learning_rate": 3.4108e-07,
+      "loss": 0.4082,
+      "step": 164800
+    },
+    {
+      "epoch": 1.6484999999999999,
+      "grad_norm": 46.557071685791016,
+      "learning_rate": 3.4088e-07,
+      "loss": 0.3833,
+      "step": 164850
+    },
+    {
+      "epoch": 1.649,
+      "grad_norm": 109.10408782958984,
+      "learning_rate": 3.4067999999999995e-07,
+      "loss": 0.3742,
+      "step": 164900
+    },
+    {
+      "epoch": 1.6495,
+      "grad_norm": 84.44685363769531,
+      "learning_rate": 3.4048e-07,
+      "loss": 0.4855,
+      "step": 164950
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.448240041732788,
+      "learning_rate": 3.4028000000000003e-07,
+      "loss": 0.5069,
+      "step": 165000
+    },
+    {
+      "epoch": 1.6505,
+      "grad_norm": 25.00313377380371,
+      "learning_rate": 3.4007999999999996e-07,
+      "loss": 0.4354,
+      "step": 165050
+    },
+    {
+      "epoch": 1.651,
+      "grad_norm": 83.84122467041016,
+      "learning_rate": 3.3988e-07,
+      "loss": 0.5135,
+      "step": 165100
+    },
+    {
+      "epoch": 1.6515,
+      "grad_norm": 0.6992330551147461,
+      "learning_rate": 3.3968e-07,
+      "loss": 0.3979,
+      "step": 165150
+    },
+    {
+      "epoch": 1.6520000000000001,
+      "grad_norm": 13.750035285949707,
+      "learning_rate": 3.3947999999999997e-07,
+      "loss": 0.4341,
+      "step": 165200
+    },
+    {
+      "epoch": 1.6524999999999999,
+      "grad_norm": 7.680423736572266,
+      "learning_rate": 3.3928e-07,
+      "loss": 0.4273,
+      "step": 165250
+    },
+    {
+      "epoch": 1.653,
+      "grad_norm": 1.2927963733673096,
+      "learning_rate": 3.3908e-07,
+      "loss": 0.3914,
+      "step": 165300
+    },
+    {
+      "epoch": 1.6535,
+      "grad_norm": 92.42204284667969,
+      "learning_rate": 3.3888e-07,
+      "loss": 0.5286,
+      "step": 165350
+    },
+    {
+      "epoch": 1.654,
+      "grad_norm": 102.65548706054688,
+      "learning_rate": 3.3867999999999996e-07,
+      "loss": 0.485,
+      "step": 165400
+    },
+    {
+      "epoch": 1.6545,
+      "grad_norm": 7.027461051940918,
+      "learning_rate": 3.3848e-07,
+      "loss": 0.4433,
+      "step": 165450
+    },
+    {
+      "epoch": 1.655,
+      "grad_norm": 74.44886779785156,
+      "learning_rate": 3.3828000000000004e-07,
+      "loss": 0.3997,
+      "step": 165500
+    },
+    {
+      "epoch": 1.6555,
+      "grad_norm": 36.282901763916016,
+      "learning_rate": 3.3807999999999997e-07,
+      "loss": 0.389,
+      "step": 165550
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 23.829561233520508,
+      "learning_rate": 3.3788e-07,
+      "loss": 0.312,
+      "step": 165600
+    },
+    {
+      "epoch": 1.6564999999999999,
+      "grad_norm": 50.19546127319336,
+      "learning_rate": 3.3768e-07,
+      "loss": 0.4511,
+      "step": 165650
+    },
+    {
+      "epoch": 1.657,
+      "grad_norm": 37.749725341796875,
+      "learning_rate": 3.3748e-07,
+      "loss": 0.3837,
+      "step": 165700
+    },
+    {
+      "epoch": 1.6575,
+      "grad_norm": 8.897858619689941,
+      "learning_rate": 3.3728e-07,
+      "loss": 0.5425,
+      "step": 165750
+    },
+    {
+      "epoch": 1.658,
+      "grad_norm": 3.789098024368286,
+      "learning_rate": 3.3708e-07,
+      "loss": 0.4618,
+      "step": 165800
+    },
+    {
+      "epoch": 1.6585,
+      "grad_norm": 73.6874771118164,
+      "learning_rate": 3.3688e-07,
+      "loss": 0.6174,
+      "step": 165850
+    },
+    {
+      "epoch": 1.659,
+      "grad_norm": 28.941118240356445,
+      "learning_rate": 3.3667999999999997e-07,
+      "loss": 0.4578,
+      "step": 165900
+    },
+    {
+      "epoch": 1.6595,
+      "grad_norm": 104.42362213134766,
+      "learning_rate": 3.3648e-07,
+      "loss": 0.4737,
+      "step": 165950
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 6.5759477615356445,
+      "learning_rate": 3.3628e-07,
+      "loss": 0.4182,
+      "step": 166000
+    },
+    {
+      "epoch": 1.6604999999999999,
+      "grad_norm": 3.6129372119903564,
+      "learning_rate": 3.3608e-07,
+      "loss": 0.3286,
+      "step": 166050
+    },
+    {
+      "epoch": 1.661,
+      "grad_norm": 32.19172668457031,
+      "learning_rate": 3.3588e-07,
+      "loss": 0.3299,
+      "step": 166100
+    },
+    {
+      "epoch": 1.6615,
+      "grad_norm": 34.54439163208008,
+      "learning_rate": 3.3567999999999995e-07,
+      "loss": 0.2821,
+      "step": 166150
+    },
+    {
+      "epoch": 1.662,
+      "grad_norm": 7.3551411628723145,
+      "learning_rate": 3.3548e-07,
+      "loss": 0.6133,
+      "step": 166200
+    },
+    {
+      "epoch": 1.6625,
+      "grad_norm": 135.5836944580078,
+      "learning_rate": 3.3528e-07,
+      "loss": 0.5549,
+      "step": 166250
+    },
+    {
+      "epoch": 1.663,
+      "grad_norm": 65.05741882324219,
+      "learning_rate": 3.3507999999999995e-07,
+      "loss": 0.5788,
+      "step": 166300
+    },
+    {
+      "epoch": 1.6635,
+      "grad_norm": 46.17664337158203,
+      "learning_rate": 3.3488e-07,
+      "loss": 0.5116,
+      "step": 166350
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 55.95429611206055,
+      "learning_rate": 3.3468e-07,
+      "loss": 0.4453,
+      "step": 166400
+    },
+    {
+      "epoch": 1.6644999999999999,
+      "grad_norm": 1.1684380769729614,
+      "learning_rate": 3.3447999999999996e-07,
+      "loss": 0.4546,
+      "step": 166450
+    },
+    {
+      "epoch": 1.665,
+      "grad_norm": 4.0728230476379395,
+      "learning_rate": 3.3428e-07,
+      "loss": 0.4625,
+      "step": 166500
+    },
+    {
+      "epoch": 1.6655,
+      "grad_norm": 14.39780330657959,
+      "learning_rate": 3.3408e-07,
+      "loss": 0.3562,
+      "step": 166550
+    },
+    {
+      "epoch": 1.666,
+      "grad_norm": 0.6815798282623291,
+      "learning_rate": 3.3387999999999997e-07,
+      "loss": 0.4597,
+      "step": 166600
+    },
+    {
+      "epoch": 1.6665,
+      "grad_norm": 95.65667724609375,
+      "learning_rate": 3.3367999999999995e-07,
+      "loss": 0.5346,
+      "step": 166650
+    },
+    {
+      "epoch": 1.667,
+      "grad_norm": 13.627656936645508,
+      "learning_rate": 3.3348e-07,
+      "loss": 0.4855,
+      "step": 166700
+    },
+    {
+      "epoch": 1.6675,
+      "grad_norm": 5.321529388427734,
+      "learning_rate": 3.3328000000000003e-07,
+      "loss": 0.401,
+      "step": 166750
+    },
+    {
+      "epoch": 1.6680000000000001,
+      "grad_norm": 9.042032241821289,
+      "learning_rate": 3.3307999999999996e-07,
+      "loss": 0.3916,
+      "step": 166800
+    },
+    {
+      "epoch": 1.6684999999999999,
+      "grad_norm": 2.1465108394622803,
+      "learning_rate": 3.3288e-07,
+      "loss": 0.559,
+      "step": 166850
+    },
+    {
+      "epoch": 1.669,
+      "grad_norm": 60.72265625,
+      "learning_rate": 3.3268e-07,
+      "loss": 0.3849,
+      "step": 166900
+    },
+    {
+      "epoch": 1.6695,
+      "grad_norm": 68.50015258789062,
+      "learning_rate": 3.3247999999999997e-07,
+      "loss": 0.4911,
+      "step": 166950
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.50178337097168,
+      "learning_rate": 3.3228e-07,
+      "loss": 0.4197,
+      "step": 167000
+    },
+    {
+      "epoch": 1.6705,
+      "grad_norm": 3.2465248107910156,
+      "learning_rate": 3.3208e-07,
+      "loss": 0.3693,
+      "step": 167050
+    },
+    {
+      "epoch": 1.671,
+      "grad_norm": 7.669544219970703,
+      "learning_rate": 3.3188e-07,
+      "loss": 0.4919,
+      "step": 167100
+    },
+    {
+      "epoch": 1.6715,
+      "grad_norm": 50.00425720214844,
+      "learning_rate": 3.3167999999999996e-07,
+      "loss": 0.4257,
+      "step": 167150
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 2.649852991104126,
+      "learning_rate": 3.3148e-07,
+      "loss": 0.43,
+      "step": 167200
+    },
+    {
+      "epoch": 1.6724999999999999,
+      "grad_norm": 70.55695343017578,
+      "learning_rate": 3.3128e-07,
+      "loss": 0.3321,
+      "step": 167250
+    },
+    {
+      "epoch": 1.673,
+      "grad_norm": 18.895328521728516,
+      "learning_rate": 3.3107999999999997e-07,
+      "loss": 0.5031,
+      "step": 167300
+    },
+    {
+      "epoch": 1.6735,
+      "grad_norm": 7.735249042510986,
+      "learning_rate": 3.3088e-07,
+      "loss": 0.501,
+      "step": 167350
+    },
+    {
+      "epoch": 1.674,
+      "grad_norm": 60.472511291503906,
+      "learning_rate": 3.3067999999999994e-07,
+      "loss": 0.3583,
+      "step": 167400
+    },
+    {
+      "epoch": 1.6745,
+      "grad_norm": 121.01581573486328,
+      "learning_rate": 3.3048e-07,
+      "loss": 0.4813,
+      "step": 167450
+    },
+    {
+      "epoch": 1.675,
+      "grad_norm": 31.92609405517578,
+      "learning_rate": 3.3028e-07,
+      "loss": 0.4163,
+      "step": 167500
+    },
+    {
+      "epoch": 1.6755,
+      "grad_norm": 20.554677963256836,
+      "learning_rate": 3.3007999999999995e-07,
+      "loss": 0.3398,
+      "step": 167550
+    },
+    {
+      "epoch": 1.6760000000000002,
+      "grad_norm": 50.82945251464844,
+      "learning_rate": 3.2988e-07,
+      "loss": 0.5493,
+      "step": 167600
+    },
+    {
+      "epoch": 1.6764999999999999,
+      "grad_norm": 13.953234672546387,
+      "learning_rate": 3.2967999999999997e-07,
+      "loss": 0.4574,
+      "step": 167650
+    },
+    {
+      "epoch": 1.677,
+      "grad_norm": 16.389623641967773,
+      "learning_rate": 3.2947999999999996e-07,
+      "loss": 0.3857,
+      "step": 167700
+    },
+    {
+      "epoch": 1.6775,
+      "grad_norm": 2.509197235107422,
+      "learning_rate": 3.2928e-07,
+      "loss": 0.3919,
+      "step": 167750
+    },
+    {
+      "epoch": 1.678,
+      "grad_norm": 4.2857666015625,
+      "learning_rate": 3.2908e-07,
+      "loss": 0.4109,
+      "step": 167800
+    },
+    {
+      "epoch": 1.6785,
+      "grad_norm": 35.18456268310547,
+      "learning_rate": 3.2887999999999996e-07,
+      "loss": 0.4017,
+      "step": 167850
+    },
+    {
+      "epoch": 1.679,
+      "grad_norm": 76.32749938964844,
+      "learning_rate": 3.2868e-07,
+      "loss": 0.3254,
+      "step": 167900
+    },
+    {
+      "epoch": 1.6795,
+      "grad_norm": 11.019416809082031,
+      "learning_rate": 3.2848e-07,
+      "loss": 0.4384,
+      "step": 167950
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 18.828954696655273,
+      "learning_rate": 3.2828e-07,
+      "loss": 0.3532,
+      "step": 168000
+    },
+    {
+      "epoch": 1.6804999999999999,
+      "grad_norm": 51.45600891113281,
+      "learning_rate": 3.2807999999999996e-07,
+      "loss": 0.4289,
+      "step": 168050
+    },
+    {
+      "epoch": 1.681,
+      "grad_norm": 38.062007904052734,
+      "learning_rate": 3.2788e-07,
+      "loss": 0.3752,
+      "step": 168100
+    },
+    {
+      "epoch": 1.6815,
+      "grad_norm": 135.77662658691406,
+      "learning_rate": 3.2768000000000003e-07,
+      "loss": 0.3286,
+      "step": 168150
+    },
+    {
+      "epoch": 1.682,
+      "grad_norm": 79.01766967773438,
+      "learning_rate": 3.2747999999999997e-07,
+      "loss": 0.4695,
+      "step": 168200
+    },
+    {
+      "epoch": 1.6825,
+      "grad_norm": 61.94942855834961,
+      "learning_rate": 3.2728e-07,
+      "loss": 0.2963,
+      "step": 168250
+    },
+    {
+      "epoch": 1.683,
+      "grad_norm": 101.182373046875,
+      "learning_rate": 3.2708e-07,
+      "loss": 0.4316,
+      "step": 168300
+    },
+    {
+      "epoch": 1.6835,
+      "grad_norm": 44.68824768066406,
+      "learning_rate": 3.2687999999999997e-07,
+      "loss": 0.2822,
+      "step": 168350
+    },
+    {
+      "epoch": 1.6840000000000002,
+      "grad_norm": 97.19388580322266,
+      "learning_rate": 3.2668e-07,
+      "loss": 0.4836,
+      "step": 168400
+    },
+    {
+      "epoch": 1.6844999999999999,
+      "grad_norm": 55.13529968261719,
+      "learning_rate": 3.2648e-07,
+      "loss": 0.2778,
+      "step": 168450
+    },
+    {
+      "epoch": 1.685,
+      "grad_norm": 7.870936870574951,
+      "learning_rate": 3.2628e-07,
+      "loss": 0.4664,
+      "step": 168500
+    },
+    {
+      "epoch": 1.6855,
+      "grad_norm": 2.3908069133758545,
+      "learning_rate": 3.2607999999999997e-07,
+      "loss": 0.5063,
+      "step": 168550
+    },
+    {
+      "epoch": 1.686,
+      "grad_norm": 4.8086466789245605,
+      "learning_rate": 3.2588e-07,
+      "loss": 0.3995,
+      "step": 168600
+    },
+    {
+      "epoch": 1.6865,
+      "grad_norm": 0.3409883677959442,
+      "learning_rate": 3.2568e-07,
+      "loss": 0.3984,
+      "step": 168650
+    },
+    {
+      "epoch": 1.687,
+      "grad_norm": 10.495270729064941,
+      "learning_rate": 3.2548e-07,
+      "loss": 0.4395,
+      "step": 168700
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 41.94511032104492,
+      "learning_rate": 3.2528e-07,
+      "loss": 0.4126,
+      "step": 168750
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 57.92393493652344,
+      "learning_rate": 3.2507999999999994e-07,
+      "loss": 0.318,
+      "step": 168800
+    },
+    {
+      "epoch": 1.6885,
+      "grad_norm": 58.26197814941406,
+      "learning_rate": 3.2488e-07,
+      "loss": 0.4174,
+      "step": 168850
+    },
+    {
+      "epoch": 1.689,
+      "grad_norm": 7.934047222137451,
+      "learning_rate": 3.2468e-07,
+      "loss": 0.4684,
+      "step": 168900
+    },
+    {
+      "epoch": 1.6895,
+      "grad_norm": 76.46321105957031,
+      "learning_rate": 3.2447999999999995e-07,
+      "loss": 0.3206,
+      "step": 168950
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 55.63337707519531,
+      "learning_rate": 3.2428e-07,
+      "loss": 0.5596,
+      "step": 169000
+    },
+    {
+      "epoch": 1.6905000000000001,
+      "grad_norm": 9.17141056060791,
+      "learning_rate": 3.2408e-07,
+      "loss": 0.3573,
+      "step": 169050
+    },
+    {
+      "epoch": 1.6909999999999998,
+      "grad_norm": 76.35599517822266,
+      "learning_rate": 3.2388e-07,
+      "loss": 0.6063,
+      "step": 169100
+    },
+    {
+      "epoch": 1.6915,
+      "grad_norm": 54.43082809448242,
+      "learning_rate": 3.2368e-07,
+      "loss": 0.4363,
+      "step": 169150
+    },
+    {
+      "epoch": 1.692,
+      "grad_norm": 25.601472854614258,
+      "learning_rate": 3.2348e-07,
+      "loss": 0.3161,
+      "step": 169200
+    },
+    {
+      "epoch": 1.6925,
+      "grad_norm": 61.37073516845703,
+      "learning_rate": 3.2328e-07,
+      "loss": 0.4411,
+      "step": 169250
+    },
+    {
+      "epoch": 1.693,
+      "grad_norm": 73.22949981689453,
+      "learning_rate": 3.2307999999999995e-07,
+      "loss": 0.4661,
+      "step": 169300
+    },
+    {
+      "epoch": 1.6935,
+      "grad_norm": 16.108476638793945,
+      "learning_rate": 3.2288e-07,
+      "loss": 0.4321,
+      "step": 169350
+    },
+    {
+      "epoch": 1.694,
+      "grad_norm": 83.31065368652344,
+      "learning_rate": 3.2268000000000003e-07,
+      "loss": 0.3823,
+      "step": 169400
+    },
+    {
+      "epoch": 1.6945000000000001,
+      "grad_norm": 1.437071681022644,
+      "learning_rate": 3.2248399999999996e-07,
+      "loss": 0.4882,
+      "step": 169450
+    },
+    {
+      "epoch": 1.6949999999999998,
+      "grad_norm": 60.666839599609375,
+      "learning_rate": 3.22284e-07,
+      "loss": 0.4287,
+      "step": 169500
+    },
+    {
+      "epoch": 1.6955,
+      "grad_norm": 52.73131561279297,
+      "learning_rate": 3.22084e-07,
+      "loss": 0.3638,
+      "step": 169550
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 67.59291076660156,
+      "learning_rate": 3.2188399999999997e-07,
+      "loss": 0.5091,
+      "step": 169600
+    },
+    {
+      "epoch": 1.6965,
+      "grad_norm": 33.913936614990234,
+      "learning_rate": 3.21684e-07,
+      "loss": 0.3616,
+      "step": 169650
+    },
+    {
+      "epoch": 1.697,
+      "grad_norm": 6.20016622543335,
+      "learning_rate": 3.21484e-07,
+      "loss": 0.4313,
+      "step": 169700
+    },
+    {
+      "epoch": 1.6975,
+      "grad_norm": 78.34620666503906,
+      "learning_rate": 3.21288e-07,
+      "loss": 0.4994,
+      "step": 169750
+    },
+    {
+      "epoch": 1.698,
+      "grad_norm": 8.514922142028809,
+      "learning_rate": 3.2108799999999997e-07,
+      "loss": 0.4833,
+      "step": 169800
+    },
+    {
+      "epoch": 1.6985000000000001,
+      "grad_norm": 43.1236572265625,
+      "learning_rate": 3.20888e-07,
+      "loss": 0.4908,
+      "step": 169850
+    },
+    {
+      "epoch": 1.6989999999999998,
+      "grad_norm": 51.2786750793457,
+      "learning_rate": 3.2068799999999994e-07,
+      "loss": 0.6363,
+      "step": 169900
+    },
+    {
+      "epoch": 1.6995,
+      "grad_norm": 102.73619842529297,
+      "learning_rate": 3.20488e-07,
+      "loss": 0.442,
+      "step": 169950
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 121.2522201538086,
+      "learning_rate": 3.20288e-07,
+      "loss": 0.5802,
+      "step": 170000
+    },
+    {
+      "epoch": 1.7005,
+      "grad_norm": 31.411584854125977,
+      "learning_rate": 3.2008799999999995e-07,
+      "loss": 0.3274,
+      "step": 170050
+    },
+    {
+      "epoch": 1.701,
+      "grad_norm": 83.65167999267578,
+      "learning_rate": 3.19888e-07,
+      "loss": 0.3685,
+      "step": 170100
+    },
+    {
+      "epoch": 1.7015,
+      "grad_norm": 11.413860321044922,
+      "learning_rate": 3.19688e-07,
+      "loss": 0.382,
+      "step": 170150
+    },
+    {
+      "epoch": 1.702,
+      "grad_norm": 18.51053237915039,
+      "learning_rate": 3.1948799999999996e-07,
+      "loss": 0.3183,
+      "step": 170200
+    },
+    {
+      "epoch": 1.7025000000000001,
+      "grad_norm": 4.223658084869385,
+      "learning_rate": 3.19288e-07,
+      "loss": 0.2734,
+      "step": 170250
+    },
+    {
+      "epoch": 1.7029999999999998,
+      "grad_norm": 4.411848545074463,
+      "learning_rate": 3.19088e-07,
+      "loss": 0.3668,
+      "step": 170300
+    },
+    {
+      "epoch": 1.7035,
+      "grad_norm": 29.715543746948242,
+      "learning_rate": 3.18888e-07,
+      "loss": 0.3997,
+      "step": 170350
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 3.227698802947998,
+      "learning_rate": 3.18688e-07,
+      "loss": 0.5644,
+      "step": 170400
+    },
+    {
+      "epoch": 1.7045,
+      "grad_norm": 41.32566833496094,
+      "learning_rate": 3.18488e-07,
+      "loss": 0.5333,
+      "step": 170450
+    },
+    {
+      "epoch": 1.705,
+      "grad_norm": 11.538256645202637,
+      "learning_rate": 3.18288e-07,
+      "loss": 0.3157,
+      "step": 170500
+    },
+    {
+      "epoch": 1.7055,
+      "grad_norm": 0.46403923630714417,
+      "learning_rate": 3.1808799999999996e-07,
+      "loss": 0.534,
+      "step": 170550
+    },
+    {
+      "epoch": 1.706,
+      "grad_norm": 18.477012634277344,
+      "learning_rate": 3.17892e-07,
+      "loss": 0.4467,
+      "step": 170600
+    },
+    {
+      "epoch": 1.7065000000000001,
+      "grad_norm": 69.50625610351562,
+      "learning_rate": 3.17692e-07,
+      "loss": 0.4261,
+      "step": 170650
+    },
+    {
+      "epoch": 1.7069999999999999,
+      "grad_norm": 73.91793060302734,
+      "learning_rate": 3.1749199999999997e-07,
+      "loss": 0.4084,
+      "step": 170700
+    },
+    {
+      "epoch": 1.7075,
+      "grad_norm": 43.193870544433594,
+      "learning_rate": 3.17292e-07,
+      "loss": 0.3532,
+      "step": 170750
+    },
+    {
+      "epoch": 1.708,
+      "grad_norm": 29.659255981445312,
+      "learning_rate": 3.17092e-07,
+      "loss": 0.4348,
+      "step": 170800
+    },
+    {
+      "epoch": 1.7085,
+      "grad_norm": 3.251600980758667,
+      "learning_rate": 3.16892e-07,
+      "loss": 0.4886,
+      "step": 170850
+    },
+    {
+      "epoch": 1.709,
+      "grad_norm": 21.75658416748047,
+      "learning_rate": 3.1669199999999996e-07,
+      "loss": 0.515,
+      "step": 170900
+    },
+    {
+      "epoch": 1.7095,
+      "grad_norm": 6.0939106941223145,
+      "learning_rate": 3.16492e-07,
+      "loss": 0.5142,
+      "step": 170950
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 94.46733856201172,
+      "learning_rate": 3.16292e-07,
+      "loss": 0.5497,
+      "step": 171000
+    },
+    {
+      "epoch": 1.7105000000000001,
+      "grad_norm": 148.20986938476562,
+      "learning_rate": 3.1609199999999997e-07,
+      "loss": 0.5326,
+      "step": 171050
+    },
+    {
+      "epoch": 1.7109999999999999,
+      "grad_norm": 25.23739242553711,
+      "learning_rate": 3.15892e-07,
+      "loss": 0.4548,
+      "step": 171100
+    },
+    {
+      "epoch": 1.7115,
+      "grad_norm": 45.993045806884766,
+      "learning_rate": 3.1569199999999994e-07,
+      "loss": 0.3861,
+      "step": 171150
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 70.52889251708984,
+      "learning_rate": 3.15492e-07,
+      "loss": 0.4945,
+      "step": 171200
+    },
+    {
+      "epoch": 1.7125,
+      "grad_norm": 0.17195826768875122,
+      "learning_rate": 3.15292e-07,
+      "loss": 0.4047,
+      "step": 171250
+    },
+    {
+      "epoch": 1.713,
+      "grad_norm": 141.29371643066406,
+      "learning_rate": 3.1509199999999995e-07,
+      "loss": 0.4854,
+      "step": 171300
+    },
+    {
+      "epoch": 1.7135,
+      "grad_norm": 79.38813781738281,
+      "learning_rate": 3.14892e-07,
+      "loss": 0.4288,
+      "step": 171350
+    },
+    {
+      "epoch": 1.714,
+      "grad_norm": 93.10710144042969,
+      "learning_rate": 3.14692e-07,
+      "loss": 0.3964,
+      "step": 171400
+    },
+    {
+      "epoch": 1.7145000000000001,
+      "grad_norm": 120.59520721435547,
+      "learning_rate": 3.1449199999999995e-07,
+      "loss": 0.3408,
+      "step": 171450
+    },
+    {
+      "epoch": 1.7149999999999999,
+      "grad_norm": 33.953556060791016,
+      "learning_rate": 3.14292e-07,
+      "loss": 0.4579,
+      "step": 171500
+    },
+    {
+      "epoch": 1.7155,
+      "grad_norm": 32.431541442871094,
+      "learning_rate": 3.14092e-07,
+      "loss": 0.348,
+      "step": 171550
+    },
+    {
+      "epoch": 1.716,
+      "grad_norm": 36.8668327331543,
+      "learning_rate": 3.13892e-07,
+      "loss": 0.4879,
+      "step": 171600
+    },
+    {
+      "epoch": 1.7165,
+      "grad_norm": 2.457643747329712,
+      "learning_rate": 3.13692e-07,
+      "loss": 0.3101,
+      "step": 171650
+    },
+    {
+      "epoch": 1.717,
+      "grad_norm": 29.928564071655273,
+      "learning_rate": 3.13496e-07,
+      "loss": 0.433,
+      "step": 171700
+    },
+    {
+      "epoch": 1.7175,
+      "grad_norm": 46.240840911865234,
+      "learning_rate": 3.1329600000000003e-07,
+      "loss": 0.4581,
+      "step": 171750
+    },
+    {
+      "epoch": 1.718,
+      "grad_norm": 13.193766593933105,
+      "learning_rate": 3.1309599999999996e-07,
+      "loss": 0.5739,
+      "step": 171800
+    },
+    {
+      "epoch": 1.7185000000000001,
+      "grad_norm": 47.43181228637695,
+      "learning_rate": 3.12896e-07,
+      "loss": 0.4211,
+      "step": 171850
+    },
+    {
+      "epoch": 1.7189999999999999,
+      "grad_norm": 50.500213623046875,
+      "learning_rate": 3.12696e-07,
+      "loss": 0.3436,
+      "step": 171900
+    },
+    {
+      "epoch": 1.7195,
+      "grad_norm": 5.8976731300354,
+      "learning_rate": 3.1249599999999997e-07,
+      "loss": 0.3893,
+      "step": 171950
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 71.6220474243164,
+      "learning_rate": 3.12296e-07,
+      "loss": 0.4529,
+      "step": 172000
+    },
+    {
+      "epoch": 1.7205,
+      "grad_norm": 32.8338508605957,
+      "learning_rate": 3.12096e-07,
+      "loss": 0.3568,
+      "step": 172050
+    },
+    {
+      "epoch": 1.721,
+      "grad_norm": 17.878786087036133,
+      "learning_rate": 3.11896e-07,
+      "loss": 0.2774,
+      "step": 172100
+    },
+    {
+      "epoch": 1.7215,
+      "grad_norm": 90.94465637207031,
+      "learning_rate": 3.1169599999999996e-07,
+      "loss": 0.3775,
+      "step": 172150
+    },
+    {
+      "epoch": 1.722,
+      "grad_norm": 101.07366180419922,
+      "learning_rate": 3.11496e-07,
+      "loss": 0.452,
+      "step": 172200
+    },
+    {
+      "epoch": 1.7225000000000001,
+      "grad_norm": 61.204246520996094,
+      "learning_rate": 3.11296e-07,
+      "loss": 0.4113,
+      "step": 172250
+    },
+    {
+      "epoch": 1.7229999999999999,
+      "grad_norm": 92.17945098876953,
+      "learning_rate": 3.1109599999999997e-07,
+      "loss": 0.4296,
+      "step": 172300
+    },
+    {
+      "epoch": 1.7235,
+      "grad_norm": 74.0315170288086,
+      "learning_rate": 3.10896e-07,
+      "loss": 0.5087,
+      "step": 172350
+    },
+    {
+      "epoch": 1.724,
+      "grad_norm": 91.9222640991211,
+      "learning_rate": 3.10696e-07,
+      "loss": 0.4429,
+      "step": 172400
+    },
+    {
+      "epoch": 1.7245,
+      "grad_norm": 33.95732498168945,
+      "learning_rate": 3.10496e-07,
+      "loss": 0.5813,
+      "step": 172450
+    },
+    {
+      "epoch": 1.725,
+      "grad_norm": 80.2164535522461,
+      "learning_rate": 3.10296e-07,
+      "loss": 0.4709,
+      "step": 172500
+    },
+    {
+      "epoch": 1.7255,
+      "grad_norm": 131.9748992919922,
+      "learning_rate": 3.1009599999999995e-07,
+      "loss": 0.4982,
+      "step": 172550
+    },
+    {
+      "epoch": 1.726,
+      "grad_norm": 1.9716410636901855,
+      "learning_rate": 3.09896e-07,
+      "loss": 0.3589,
+      "step": 172600
+    },
+    {
+      "epoch": 1.7265000000000001,
+      "grad_norm": 76.4041519165039,
+      "learning_rate": 3.09696e-07,
+      "loss": 0.4375,
+      "step": 172650
+    },
+    {
+      "epoch": 1.7269999999999999,
+      "grad_norm": 10.473490715026855,
+      "learning_rate": 3.0949599999999995e-07,
+      "loss": 0.3493,
+      "step": 172700
+    },
+    {
+      "epoch": 1.7275,
+      "grad_norm": 44.0904426574707,
+      "learning_rate": 3.09296e-07,
+      "loss": 0.4193,
+      "step": 172750
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 37.536067962646484,
+      "learning_rate": 3.09096e-07,
+      "loss": 0.4956,
+      "step": 172800
+    },
+    {
+      "epoch": 1.7285,
+      "grad_norm": 141.21176147460938,
+      "learning_rate": 3.08896e-07,
+      "loss": 0.3959,
+      "step": 172850
+    },
+    {
+      "epoch": 1.729,
+      "grad_norm": 15.123921394348145,
+      "learning_rate": 3.08696e-07,
+      "loss": 0.4004,
+      "step": 172900
+    },
+    {
+      "epoch": 1.7295,
+      "grad_norm": 38.241249084472656,
+      "learning_rate": 3.08496e-07,
+      "loss": 0.5414,
+      "step": 172950
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 87.40081787109375,
+      "learning_rate": 3.08296e-07,
+      "loss": 0.5123,
+      "step": 173000
+    },
+    {
+      "epoch": 1.7305000000000001,
+      "grad_norm": 17.00108528137207,
+      "learning_rate": 3.0809599999999996e-07,
+      "loss": 0.4696,
+      "step": 173050
+    },
+    {
+      "epoch": 1.7309999999999999,
+      "grad_norm": 5.777243614196777,
+      "learning_rate": 3.07896e-07,
+      "loss": 0.479,
+      "step": 173100
+    },
+    {
+      "epoch": 1.7315,
+      "grad_norm": 92.2618179321289,
+      "learning_rate": 3.0769600000000003e-07,
+      "loss": 0.3697,
+      "step": 173150
+    },
+    {
+      "epoch": 1.732,
+      "grad_norm": 115.76209259033203,
+      "learning_rate": 3.0749599999999996e-07,
+      "loss": 0.5014,
+      "step": 173200
+    },
+    {
+      "epoch": 1.7325,
+      "grad_norm": 10.42154598236084,
+      "learning_rate": 3.07296e-07,
+      "loss": 0.521,
+      "step": 173250
+    },
+    {
+      "epoch": 1.733,
+      "grad_norm": 48.72549819946289,
+      "learning_rate": 3.07096e-07,
+      "loss": 0.5236,
+      "step": 173300
+    },
+    {
+      "epoch": 1.7335,
+      "grad_norm": 43.980899810791016,
+      "learning_rate": 3.0689599999999997e-07,
+      "loss": 0.5085,
+      "step": 173350
+    },
+    {
+      "epoch": 1.734,
+      "grad_norm": 7.443145275115967,
+      "learning_rate": 3.06696e-07,
+      "loss": 0.4113,
+      "step": 173400
+    },
+    {
+      "epoch": 1.7345000000000002,
+      "grad_norm": 9.301448822021484,
+      "learning_rate": 3.06496e-07,
+      "loss": 0.4126,
+      "step": 173450
+    },
+    {
+      "epoch": 1.7349999999999999,
+      "grad_norm": 85.95564270019531,
+      "learning_rate": 3.06296e-07,
+      "loss": 0.6378,
+      "step": 173500
+    },
+    {
+      "epoch": 1.7355,
+      "grad_norm": 20.996845245361328,
+      "learning_rate": 3.0609599999999996e-07,
+      "loss": 0.6383,
+      "step": 173550
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 50.388126373291016,
+      "learning_rate": 3.05896e-07,
+      "loss": 0.3279,
+      "step": 173600
+    },
+    {
+      "epoch": 1.7365,
+      "grad_norm": 0.6028560400009155,
+      "learning_rate": 3.05696e-07,
+      "loss": 0.4151,
+      "step": 173650
+    },
+    {
+      "epoch": 1.737,
+      "grad_norm": 31.13410186767578,
+      "learning_rate": 3.0549599999999997e-07,
+      "loss": 0.423,
+      "step": 173700
+    },
+    {
+      "epoch": 1.7375,
+      "grad_norm": 72.78054809570312,
+      "learning_rate": 3.05296e-07,
+      "loss": 0.344,
+      "step": 173750
+    },
+    {
+      "epoch": 1.738,
+      "grad_norm": 25.89346694946289,
+      "learning_rate": 3.0509599999999994e-07,
+      "loss": 0.6206,
+      "step": 173800
+    },
+    {
+      "epoch": 1.7385000000000002,
+      "grad_norm": 48.15966033935547,
+      "learning_rate": 3.04896e-07,
+      "loss": 0.3109,
+      "step": 173850
+    },
+    {
+      "epoch": 1.7389999999999999,
+      "grad_norm": 45.8206787109375,
+      "learning_rate": 3.04696e-07,
+      "loss": 0.5,
+      "step": 173900
+    },
+    {
+      "epoch": 1.7395,
+      "grad_norm": 9.931618690490723,
+      "learning_rate": 3.0449599999999995e-07,
+      "loss": 0.4324,
+      "step": 173950
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 107.68231964111328,
+      "learning_rate": 3.04296e-07,
+      "loss": 0.4264,
+      "step": 174000
+    },
+    {
+      "epoch": 1.7405,
+      "grad_norm": 23.903282165527344,
+      "learning_rate": 3.0409599999999997e-07,
+      "loss": 0.4645,
+      "step": 174050
+    },
+    {
+      "epoch": 1.741,
+      "grad_norm": 60.52227783203125,
+      "learning_rate": 3.03896e-07,
+      "loss": 0.4347,
+      "step": 174100
+    },
+    {
+      "epoch": 1.7415,
+      "grad_norm": 3.167391300201416,
+      "learning_rate": 3.03696e-07,
+      "loss": 0.354,
+      "step": 174150
+    },
+    {
+      "epoch": 1.742,
+      "grad_norm": 72.62369537353516,
+      "learning_rate": 3.03496e-07,
+      "loss": 0.4243,
+      "step": 174200
+    },
+    {
+      "epoch": 1.7425000000000002,
+      "grad_norm": 51.68940734863281,
+      "learning_rate": 3.03296e-07,
+      "loss": 0.4351,
+      "step": 174250
+    },
+    {
+      "epoch": 1.7429999999999999,
+      "grad_norm": 93.96479797363281,
+      "learning_rate": 3.0309599999999995e-07,
+      "loss": 0.3287,
+      "step": 174300
+    },
+    {
+      "epoch": 1.7435,
+      "grad_norm": 151.80270385742188,
+      "learning_rate": 3.02896e-07,
+      "loss": 0.3704,
+      "step": 174350
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 6.188260555267334,
+      "learning_rate": 3.02696e-07,
+      "loss": 0.5232,
+      "step": 174400
+    },
+    {
+      "epoch": 1.7445,
+      "grad_norm": 19.85384178161621,
+      "learning_rate": 3.0249599999999996e-07,
+      "loss": 0.4082,
+      "step": 174450
+    },
+    {
+      "epoch": 1.745,
+      "grad_norm": 2.640712022781372,
+      "learning_rate": 3.02296e-07,
+      "loss": 0.3756,
+      "step": 174500
+    },
+    {
+      "epoch": 1.7455,
+      "grad_norm": 20.137554168701172,
+      "learning_rate": 3.02096e-07,
+      "loss": 0.459,
+      "step": 174550
+    },
+    {
+      "epoch": 1.746,
+      "grad_norm": 60.7595100402832,
+      "learning_rate": 3.0189599999999997e-07,
+      "loss": 0.4374,
+      "step": 174600
+    },
+    {
+      "epoch": 1.7465000000000002,
+      "grad_norm": 51.217464447021484,
+      "learning_rate": 3.01696e-07,
+      "loss": 0.4513,
+      "step": 174650
+    },
+    {
+      "epoch": 1.7469999999999999,
+      "grad_norm": 7.342816352844238,
+      "learning_rate": 3.01496e-07,
+      "loss": 0.448,
+      "step": 174700
+    },
+    {
+      "epoch": 1.7475,
+      "grad_norm": 1.492277979850769,
+      "learning_rate": 3.01296e-07,
+      "loss": 0.4433,
+      "step": 174750
+    },
+    {
+      "epoch": 1.748,
+      "grad_norm": 75.10221099853516,
+      "learning_rate": 3.0109599999999996e-07,
+      "loss": 0.2593,
+      "step": 174800
+    },
+    {
+      "epoch": 1.7485,
+      "grad_norm": 6.8189616203308105,
+      "learning_rate": 3.00896e-07,
+      "loss": 0.4845,
+      "step": 174850
+    },
+    {
+      "epoch": 1.749,
+      "grad_norm": 33.18309783935547,
+      "learning_rate": 3.00696e-07,
+      "loss": 0.3589,
+      "step": 174900
+    },
+    {
+      "epoch": 1.7495,
+      "grad_norm": 149.52420043945312,
+      "learning_rate": 3.0049599999999997e-07,
+      "loss": 0.3238,
+      "step": 174950
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 33.072242736816406,
+      "learning_rate": 3.00296e-07,
+      "loss": 0.5178,
+      "step": 175000
+    },
+    {
+      "epoch": 1.7505,
+      "grad_norm": 19.933521270751953,
+      "learning_rate": 3.0009599999999994e-07,
+      "loss": 0.4674,
+      "step": 175050
+    },
+    {
+      "epoch": 1.751,
+      "grad_norm": 14.122361183166504,
+      "learning_rate": 2.99896e-07,
+      "loss": 0.3734,
+      "step": 175100
+    },
+    {
+      "epoch": 1.7515,
+      "grad_norm": 0.9820132255554199,
+      "learning_rate": 2.99696e-07,
+      "loss": 0.3835,
+      "step": 175150
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 0.1489812433719635,
+      "learning_rate": 2.99496e-07,
+      "loss": 0.6293,
+      "step": 175200
+    },
+    {
+      "epoch": 1.7525,
+      "grad_norm": 2.5810625553131104,
+      "learning_rate": 2.99296e-07,
+      "loss": 0.335,
+      "step": 175250
+    },
+    {
+      "epoch": 1.7530000000000001,
+      "grad_norm": 4.589435577392578,
+      "learning_rate": 2.9909599999999997e-07,
+      "loss": 0.5491,
+      "step": 175300
+    },
+    {
+      "epoch": 1.7534999999999998,
+      "grad_norm": 35.78123474121094,
+      "learning_rate": 2.98896e-07,
+      "loss": 0.3732,
+      "step": 175350
+    },
+    {
+      "epoch": 1.754,
+      "grad_norm": 7.164456844329834,
+      "learning_rate": 2.98696e-07,
+      "loss": 0.418,
+      "step": 175400
+    },
+    {
+      "epoch": 1.7545,
+      "grad_norm": 102.37982177734375,
+      "learning_rate": 2.98496e-07,
+      "loss": 0.5153,
+      "step": 175450
+    },
+    {
+      "epoch": 1.755,
+      "grad_norm": 1.655551552772522,
+      "learning_rate": 2.98296e-07,
+      "loss": 0.4579,
+      "step": 175500
+    },
+    {
+      "epoch": 1.7555,
+      "grad_norm": 11.854036331176758,
+      "learning_rate": 2.98096e-07,
+      "loss": 0.5165,
+      "step": 175550
+    },
+    {
+      "epoch": 1.756,
+      "grad_norm": 5.970990180969238,
+      "learning_rate": 2.97896e-07,
+      "loss": 0.3749,
+      "step": 175600
+    },
+    {
+      "epoch": 1.7565,
+      "grad_norm": 40.14838790893555,
+      "learning_rate": 2.97696e-07,
+      "loss": 0.4894,
+      "step": 175650
+    },
+    {
+      "epoch": 1.7570000000000001,
+      "grad_norm": 6.2947187423706055,
+      "learning_rate": 2.9749999999999996e-07,
+      "loss": 0.3435,
+      "step": 175700
+    },
+    {
+      "epoch": 1.7574999999999998,
+      "grad_norm": 72.81707000732422,
+      "learning_rate": 2.973e-07,
+      "loss": 0.5341,
+      "step": 175750
+    },
+    {
+      "epoch": 1.758,
+      "grad_norm": 23.278066635131836,
+      "learning_rate": 2.971e-07,
+      "loss": 0.3967,
+      "step": 175800
+    },
+    {
+      "epoch": 1.7585,
+      "grad_norm": 75.2026138305664,
+      "learning_rate": 2.9689999999999997e-07,
+      "loss": 0.4016,
+      "step": 175850
+    },
+    {
+      "epoch": 1.759,
+      "grad_norm": 25.339069366455078,
+      "learning_rate": 2.967e-07,
+      "loss": 0.2943,
+      "step": 175900
+    },
+    {
+      "epoch": 1.7595,
+      "grad_norm": 44.909751892089844,
+      "learning_rate": 2.965e-07,
+      "loss": 0.3533,
+      "step": 175950
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 60.30315017700195,
+      "learning_rate": 2.9629999999999997e-07,
+      "loss": 0.5381,
+      "step": 176000
+    },
+    {
+      "epoch": 1.7605,
+      "grad_norm": 105.79608917236328,
+      "learning_rate": 2.9609999999999996e-07,
+      "loss": 0.6541,
+      "step": 176050
+    },
+    {
+      "epoch": 1.7610000000000001,
+      "grad_norm": 74.8968505859375,
+      "learning_rate": 2.959e-07,
+      "loss": 0.3239,
+      "step": 176100
+    },
+    {
+      "epoch": 1.7614999999999998,
+      "grad_norm": 2.4534757137298584,
+      "learning_rate": 2.957e-07,
+      "loss": 0.3273,
+      "step": 176150
+    },
+    {
+      "epoch": 1.762,
+      "grad_norm": 35.73273468017578,
+      "learning_rate": 2.9549999999999997e-07,
+      "loss": 0.4744,
+      "step": 176200
+    },
+    {
+      "epoch": 1.7625,
+      "grad_norm": 33.48553466796875,
+      "learning_rate": 2.953e-07,
+      "loss": 0.4624,
+      "step": 176250
+    },
+    {
+      "epoch": 1.763,
+      "grad_norm": 52.17403793334961,
+      "learning_rate": 2.9509999999999994e-07,
+      "loss": 0.3885,
+      "step": 176300
+    },
+    {
+      "epoch": 1.7635,
+      "grad_norm": 103.1676025390625,
+      "learning_rate": 2.949e-07,
+      "loss": 0.5388,
+      "step": 176350
+    },
+    {
+      "epoch": 1.764,
+      "grad_norm": 84.96452331542969,
+      "learning_rate": 2.947e-07,
+      "loss": 0.4278,
+      "step": 176400
+    },
+    {
+      "epoch": 1.7645,
+      "grad_norm": 1.080453634262085,
+      "learning_rate": 2.945e-07,
+      "loss": 0.5096,
+      "step": 176450
+    },
+    {
+      "epoch": 1.7650000000000001,
+      "grad_norm": 1.111628532409668,
+      "learning_rate": 2.943e-07,
+      "loss": 0.3848,
+      "step": 176500
+    },
+    {
+      "epoch": 1.7654999999999998,
+      "grad_norm": 31.323686599731445,
+      "learning_rate": 2.9409999999999997e-07,
+      "loss": 0.3379,
+      "step": 176550
+    },
+    {
+      "epoch": 1.766,
+      "grad_norm": 42.48777389526367,
+      "learning_rate": 2.939e-07,
+      "loss": 0.3058,
+      "step": 176600
+    },
+    {
+      "epoch": 1.7665,
+      "grad_norm": 7.140532970428467,
+      "learning_rate": 2.937e-07,
+      "loss": 0.3894,
+      "step": 176650
+    },
+    {
+      "epoch": 1.767,
+      "grad_norm": 18.32383918762207,
+      "learning_rate": 2.935e-07,
+      "loss": 0.2754,
+      "step": 176700
+    },
+    {
+      "epoch": 1.7675,
+      "grad_norm": 0.7897120118141174,
+      "learning_rate": 2.933e-07,
+      "loss": 0.5175,
+      "step": 176750
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 121.51934051513672,
+      "learning_rate": 2.931e-07,
+      "loss": 0.4686,
+      "step": 176800
+    },
+    {
+      "epoch": 1.7685,
+      "grad_norm": 0.11282049119472504,
+      "learning_rate": 2.929e-07,
+      "loss": 0.4456,
+      "step": 176850
+    },
+    {
+      "epoch": 1.7690000000000001,
+      "grad_norm": 19.87791633605957,
+      "learning_rate": 2.927e-07,
+      "loss": 0.3768,
+      "step": 176900
+    },
+    {
+      "epoch": 1.7694999999999999,
+      "grad_norm": 0.5945930480957031,
+      "learning_rate": 2.9249999999999995e-07,
+      "loss": 0.5334,
+      "step": 176950
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.010547161102295,
+      "learning_rate": 2.923e-07,
+      "loss": 0.4443,
+      "step": 177000
+    },
+    {
+      "epoch": 1.7705,
+      "grad_norm": 47.688934326171875,
+      "learning_rate": 2.9210000000000003e-07,
+      "loss": 0.4077,
+      "step": 177050
+    },
+    {
+      "epoch": 1.771,
+      "grad_norm": 60.86616516113281,
+      "learning_rate": 2.9189999999999996e-07,
+      "loss": 0.4187,
+      "step": 177100
+    },
+    {
+      "epoch": 1.7715,
+      "grad_norm": 19.05134391784668,
+      "learning_rate": 2.917e-07,
+      "loss": 0.4585,
+      "step": 177150
+    },
+    {
+      "epoch": 1.772,
+      "grad_norm": 41.00670623779297,
+      "learning_rate": 2.915e-07,
+      "loss": 0.3587,
+      "step": 177200
+    },
+    {
+      "epoch": 1.7725,
+      "grad_norm": 4.250216007232666,
+      "learning_rate": 2.9129999999999997e-07,
+      "loss": 0.4831,
+      "step": 177250
+    },
+    {
+      "epoch": 1.7730000000000001,
+      "grad_norm": 53.11735534667969,
+      "learning_rate": 2.911e-07,
+      "loss": 0.378,
+      "step": 177300
+    },
+    {
+      "epoch": 1.7734999999999999,
+      "grad_norm": 79.12381744384766,
+      "learning_rate": 2.909e-07,
+      "loss": 0.3421,
+      "step": 177350
+    },
+    {
+      "epoch": 1.774,
+      "grad_norm": 123.16691589355469,
+      "learning_rate": 2.907e-07,
+      "loss": 0.4126,
+      "step": 177400
+    },
+    {
+      "epoch": 1.7745,
+      "grad_norm": 83.73424530029297,
+      "learning_rate": 2.9049999999999996e-07,
+      "loss": 0.4596,
+      "step": 177450
+    },
+    {
+      "epoch": 1.775,
+      "grad_norm": 47.7365608215332,
+      "learning_rate": 2.903e-07,
+      "loss": 0.3569,
+      "step": 177500
+    },
+    {
+      "epoch": 1.7755,
+      "grad_norm": 18.262990951538086,
+      "learning_rate": 2.9010000000000004e-07,
+      "loss": 0.3546,
+      "step": 177550
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 25.71767807006836,
+      "learning_rate": 2.8989999999999997e-07,
+      "loss": 0.5118,
+      "step": 177600
+    },
+    {
+      "epoch": 1.7765,
+      "grad_norm": 43.59213638305664,
+      "learning_rate": 2.897e-07,
+      "loss": 0.4437,
+      "step": 177650
+    },
+    {
+      "epoch": 1.7770000000000001,
+      "grad_norm": 3.3139374256134033,
+      "learning_rate": 2.895e-07,
+      "loss": 0.4414,
+      "step": 177700
+    },
+    {
+      "epoch": 1.7774999999999999,
+      "grad_norm": 21.859750747680664,
+      "learning_rate": 2.893e-07,
+      "loss": 0.5413,
+      "step": 177750
+    },
+    {
+      "epoch": 1.778,
+      "grad_norm": 63.0023307800293,
+      "learning_rate": 2.891e-07,
+      "loss": 0.3183,
+      "step": 177800
+    },
+    {
+      "epoch": 1.7785,
+      "grad_norm": 7.4691267013549805,
+      "learning_rate": 2.889e-07,
+      "loss": 0.2339,
+      "step": 177850
+    },
+    {
+      "epoch": 1.779,
+      "grad_norm": 51.13058090209961,
+      "learning_rate": 2.887e-07,
+      "loss": 0.5413,
+      "step": 177900
+    },
+    {
+      "epoch": 1.7795,
+      "grad_norm": 65.8671875,
+      "learning_rate": 2.8849999999999997e-07,
+      "loss": 0.4583,
+      "step": 177950
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.142685890197754,
+      "learning_rate": 2.883e-07,
+      "loss": 0.4431,
+      "step": 178000
+    },
+    {
+      "epoch": 1.7805,
+      "grad_norm": 26.582061767578125,
+      "learning_rate": 2.881e-07,
+      "loss": 0.4004,
+      "step": 178050
+    },
+    {
+      "epoch": 1.7810000000000001,
+      "grad_norm": 73.92897033691406,
+      "learning_rate": 2.879e-07,
+      "loss": 0.5293,
+      "step": 178100
+    },
+    {
+      "epoch": 1.7814999999999999,
+      "grad_norm": 69.2878646850586,
+      "learning_rate": 2.877e-07,
+      "loss": 0.5733,
+      "step": 178150
+    },
+    {
+      "epoch": 1.782,
+      "grad_norm": 99.10835266113281,
+      "learning_rate": 2.8749999999999995e-07,
+      "loss": 0.5006,
+      "step": 178200
+    },
+    {
+      "epoch": 1.7825,
+      "grad_norm": 78.11844635009766,
+      "learning_rate": 2.873e-07,
+      "loss": 0.3794,
+      "step": 178250
+    },
+    {
+      "epoch": 1.783,
+      "grad_norm": 100.39323425292969,
+      "learning_rate": 2.871e-07,
+      "loss": 0.4381,
+      "step": 178300
+    },
+    {
+      "epoch": 1.7835,
+      "grad_norm": 19.325456619262695,
+      "learning_rate": 2.8689999999999996e-07,
+      "loss": 0.4126,
+      "step": 178350
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 3.960777759552002,
+      "learning_rate": 2.867e-07,
+      "loss": 0.4078,
+      "step": 178400
+    },
+    {
+      "epoch": 1.7845,
+      "grad_norm": 59.112300872802734,
+      "learning_rate": 2.865e-07,
+      "loss": 0.3978,
+      "step": 178450
+    },
+    {
+      "epoch": 1.7850000000000001,
+      "grad_norm": 17.307994842529297,
+      "learning_rate": 2.8629999999999996e-07,
+      "loss": 0.4808,
+      "step": 178500
+    },
+    {
+      "epoch": 1.7854999999999999,
+      "grad_norm": 33.76523971557617,
+      "learning_rate": 2.861e-07,
+      "loss": 0.4348,
+      "step": 178550
+    },
+    {
+      "epoch": 1.786,
+      "grad_norm": 80.96756744384766,
+      "learning_rate": 2.859e-07,
+      "loss": 0.4833,
+      "step": 178600
+    },
+    {
+      "epoch": 1.7865,
+      "grad_norm": 48.94674301147461,
+      "learning_rate": 2.8569999999999997e-07,
+      "loss": 0.3485,
+      "step": 178650
+    },
+    {
+      "epoch": 1.787,
+      "grad_norm": 72.22645568847656,
+      "learning_rate": 2.8549999999999996e-07,
+      "loss": 0.4025,
+      "step": 178700
+    },
+    {
+      "epoch": 1.7875,
+      "grad_norm": 4.155689716339111,
+      "learning_rate": 2.853e-07,
+      "loss": 0.405,
+      "step": 178750
+    },
+    {
+      "epoch": 1.788,
+      "grad_norm": 0.55622398853302,
+      "learning_rate": 2.8510000000000003e-07,
+      "loss": 0.5012,
+      "step": 178800
+    },
+    {
+      "epoch": 1.7885,
+      "grad_norm": 67.4764404296875,
+      "learning_rate": 2.8489999999999996e-07,
+      "loss": 0.3643,
+      "step": 178850
+    },
+    {
+      "epoch": 1.7890000000000001,
+      "grad_norm": 85.40705108642578,
+      "learning_rate": 2.847e-07,
+      "loss": 0.5223,
+      "step": 178900
+    },
+    {
+      "epoch": 1.7894999999999999,
+      "grad_norm": 85.30449676513672,
+      "learning_rate": 2.845e-07,
+      "loss": 0.4813,
+      "step": 178950
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.022859573364258,
+      "learning_rate": 2.8429999999999997e-07,
+      "loss": 0.4518,
+      "step": 179000
+    },
+    {
+      "epoch": 1.7905,
+      "grad_norm": 53.33081817626953,
+      "learning_rate": 2.841e-07,
+      "loss": 0.4076,
+      "step": 179050
+    },
+    {
+      "epoch": 1.791,
+      "grad_norm": 1.9716160297393799,
+      "learning_rate": 2.839e-07,
+      "loss": 0.4666,
+      "step": 179100
+    },
+    {
+      "epoch": 1.7915,
+      "grad_norm": 59.22763442993164,
+      "learning_rate": 2.837e-07,
+      "loss": 0.5778,
+      "step": 179150
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 7.315980911254883,
+      "learning_rate": 2.8349999999999996e-07,
+      "loss": 0.4156,
+      "step": 179200
+    },
+    {
+      "epoch": 1.7925,
+      "grad_norm": 49.91332244873047,
+      "learning_rate": 2.833e-07,
+      "loss": 0.4988,
+      "step": 179250
+    },
+    {
+      "epoch": 1.7930000000000001,
+      "grad_norm": 65.8807144165039,
+      "learning_rate": 2.831e-07,
+      "loss": 0.5433,
+      "step": 179300
+    },
+    {
+      "epoch": 1.7934999999999999,
+      "grad_norm": 103.30567169189453,
+      "learning_rate": 2.8289999999999997e-07,
+      "loss": 0.5536,
+      "step": 179350
+    },
+    {
+      "epoch": 1.794,
+      "grad_norm": 86.77356719970703,
+      "learning_rate": 2.827e-07,
+      "loss": 0.5217,
+      "step": 179400
+    },
+    {
+      "epoch": 1.7945,
+      "grad_norm": 96.92404174804688,
+      "learning_rate": 2.8249999999999994e-07,
+      "loss": 0.4565,
+      "step": 179450
+    },
+    {
+      "epoch": 1.795,
+      "grad_norm": 14.877604484558105,
+      "learning_rate": 2.823e-07,
+      "loss": 0.4828,
+      "step": 179500
+    },
+    {
+      "epoch": 1.7955,
+      "grad_norm": 24.75067901611328,
+      "learning_rate": 2.821e-07,
+      "loss": 0.4715,
+      "step": 179550
+    },
+    {
+      "epoch": 1.796,
+      "grad_norm": 45.08053207397461,
+      "learning_rate": 2.8189999999999995e-07,
+      "loss": 0.4347,
+      "step": 179600
+    },
+    {
+      "epoch": 1.7965,
+      "grad_norm": 1.8589259386062622,
+      "learning_rate": 2.817e-07,
+      "loss": 0.4739,
+      "step": 179650
+    },
+    {
+      "epoch": 1.7970000000000002,
+      "grad_norm": 128.76307678222656,
+      "learning_rate": 2.8149999999999997e-07,
+      "loss": 0.4551,
+      "step": 179700
+    },
+    {
+      "epoch": 1.7974999999999999,
+      "grad_norm": 25.413040161132812,
+      "learning_rate": 2.8129999999999996e-07,
+      "loss": 0.335,
+      "step": 179750
+    },
+    {
+      "epoch": 1.798,
+      "grad_norm": 74.6392593383789,
+      "learning_rate": 2.811e-07,
+      "loss": 0.4635,
+      "step": 179800
+    },
+    {
+      "epoch": 1.7985,
+      "grad_norm": 69.69477844238281,
+      "learning_rate": 2.809e-07,
+      "loss": 0.4422,
+      "step": 179850
+    },
+    {
+      "epoch": 1.799,
+      "grad_norm": 12.26147747039795,
+      "learning_rate": 2.807e-07,
+      "loss": 0.4213,
+      "step": 179900
+    },
+    {
+      "epoch": 1.7995,
+      "grad_norm": 90.06591033935547,
+      "learning_rate": 2.805e-07,
+      "loss": 0.3711,
+      "step": 179950
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.045799255371094,
+      "learning_rate": 2.803e-07,
+      "loss": 0.3808,
+      "step": 180000
+    },
+    {
+      "epoch": 1.8005,
+      "grad_norm": 42.867767333984375,
+      "learning_rate": 2.8010000000000003e-07,
+      "loss": 0.4572,
+      "step": 180050
+    },
+    {
+      "epoch": 1.8010000000000002,
+      "grad_norm": 63.34716033935547,
+      "learning_rate": 2.7989999999999996e-07,
+      "loss": 0.5186,
+      "step": 180100
+    },
+    {
+      "epoch": 1.8014999999999999,
+      "grad_norm": 56.33584976196289,
+      "learning_rate": 2.797e-07,
+      "loss": 0.5546,
+      "step": 180150
+    },
+    {
+      "epoch": 1.802,
+      "grad_norm": 81.60842895507812,
+      "learning_rate": 2.7950000000000003e-07,
+      "loss": 0.5097,
+      "step": 180200
+    },
+    {
+      "epoch": 1.8025,
+      "grad_norm": 12.385924339294434,
+      "learning_rate": 2.7929999999999997e-07,
+      "loss": 0.3856,
+      "step": 180250
+    },
+    {
+      "epoch": 1.803,
+      "grad_norm": 54.57160186767578,
+      "learning_rate": 2.791e-07,
+      "loss": 0.398,
+      "step": 180300
+    },
+    {
+      "epoch": 1.8035,
+      "grad_norm": 75.23123168945312,
+      "learning_rate": 2.789e-07,
+      "loss": 0.3756,
+      "step": 180350
+    },
+    {
+      "epoch": 1.804,
+      "grad_norm": 102.83467102050781,
+      "learning_rate": 2.787e-07,
+      "loss": 0.3975,
+      "step": 180400
+    },
+    {
+      "epoch": 1.8045,
+      "grad_norm": 150.63693237304688,
+      "learning_rate": 2.785e-07,
+      "loss": 0.2832,
+      "step": 180450
+    },
+    {
+      "epoch": 1.8050000000000002,
+      "grad_norm": 46.95927429199219,
+      "learning_rate": 2.783e-07,
+      "loss": 0.3984,
+      "step": 180500
+    },
+    {
+      "epoch": 1.8054999999999999,
+      "grad_norm": 103.29827117919922,
+      "learning_rate": 2.781e-07,
+      "loss": 0.3903,
+      "step": 180550
+    },
+    {
+      "epoch": 1.806,
+      "grad_norm": 6.826172351837158,
+      "learning_rate": 2.7789999999999997e-07,
+      "loss": 0.5537,
+      "step": 180600
+    },
+    {
+      "epoch": 1.8065,
+      "grad_norm": 71.7900161743164,
+      "learning_rate": 2.777e-07,
+      "loss": 0.421,
+      "step": 180650
+    },
+    {
+      "epoch": 1.807,
+      "grad_norm": 78.16989135742188,
+      "learning_rate": 2.775e-07,
+      "loss": 0.4313,
+      "step": 180700
+    },
+    {
+      "epoch": 1.8075,
+      "grad_norm": 64.42400360107422,
+      "learning_rate": 2.773e-07,
+      "loss": 0.4864,
+      "step": 180750
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 27.222681045532227,
+      "learning_rate": 2.771e-07,
+      "loss": 0.3927,
+      "step": 180800
+    },
+    {
+      "epoch": 1.8085,
+      "grad_norm": 72.81607818603516,
+      "learning_rate": 2.7689999999999995e-07,
+      "loss": 0.4236,
+      "step": 180850
+    },
+    {
+      "epoch": 1.8090000000000002,
+      "grad_norm": 119.21035766601562,
+      "learning_rate": 2.767e-07,
+      "loss": 0.5258,
+      "step": 180900
+    },
+    {
+      "epoch": 1.8094999999999999,
+      "grad_norm": 93.99649810791016,
+      "learning_rate": 2.765e-07,
+      "loss": 0.425,
+      "step": 180950
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 81.92257690429688,
+      "learning_rate": 2.7629999999999995e-07,
+      "loss": 0.4852,
+      "step": 181000
+    },
+    {
+      "epoch": 1.8105,
+      "grad_norm": 67.70683288574219,
+      "learning_rate": 2.761e-07,
+      "loss": 0.4518,
+      "step": 181050
+    },
+    {
+      "epoch": 1.811,
+      "grad_norm": 8.245046615600586,
+      "learning_rate": 2.759e-07,
+      "loss": 0.4221,
+      "step": 181100
+    },
+    {
+      "epoch": 1.8115,
+      "grad_norm": 0.5864673852920532,
+      "learning_rate": 2.757e-07,
+      "loss": 0.3862,
+      "step": 181150
+    },
+    {
+      "epoch": 1.812,
+      "grad_norm": 103.44923400878906,
+      "learning_rate": 2.755e-07,
+      "loss": 0.4913,
+      "step": 181200
+    },
+    {
+      "epoch": 1.8125,
+      "grad_norm": 0.6303176879882812,
+      "learning_rate": 2.753e-07,
+      "loss": 0.411,
+      "step": 181250
+    },
+    {
+      "epoch": 1.813,
+      "grad_norm": 50.171905517578125,
+      "learning_rate": 2.751e-07,
+      "loss": 0.3466,
+      "step": 181300
+    },
+    {
+      "epoch": 1.8135,
+      "grad_norm": 31.094560623168945,
+      "learning_rate": 2.7489999999999995e-07,
+      "loss": 0.4811,
+      "step": 181350
+    },
+    {
+      "epoch": 1.814,
+      "grad_norm": 3.8934226036071777,
+      "learning_rate": 2.747e-07,
+      "loss": 0.3938,
+      "step": 181400
+    },
+    {
+      "epoch": 1.8145,
+      "grad_norm": 37.404117584228516,
+      "learning_rate": 2.7450000000000003e-07,
+      "loss": 0.3969,
+      "step": 181450
+    },
+    {
+      "epoch": 1.815,
+      "grad_norm": 13.522012710571289,
+      "learning_rate": 2.7429999999999996e-07,
+      "loss": 0.4785,
+      "step": 181500
+    },
+    {
+      "epoch": 1.8155000000000001,
+      "grad_norm": 14.331076622009277,
+      "learning_rate": 2.741e-07,
+      "loss": 0.3068,
+      "step": 181550
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 26.77286148071289,
+      "learning_rate": 2.739e-07,
+      "loss": 0.337,
+      "step": 181600
+    },
+    {
+      "epoch": 1.8165,
+      "grad_norm": 48.4984245300293,
+      "learning_rate": 2.7369999999999997e-07,
+      "loss": 0.3276,
+      "step": 181650
+    },
+    {
+      "epoch": 1.817,
+      "grad_norm": 60.87279510498047,
+      "learning_rate": 2.735e-07,
+      "loss": 0.4779,
+      "step": 181700
+    },
+    {
+      "epoch": 1.8175,
+      "grad_norm": 51.6512336730957,
+      "learning_rate": 2.733e-07,
+      "loss": 0.3513,
+      "step": 181750
+    },
+    {
+      "epoch": 1.818,
+      "grad_norm": 6.495523929595947,
+      "learning_rate": 2.731e-07,
+      "loss": 0.5444,
+      "step": 181800
+    },
+    {
+      "epoch": 1.8185,
+      "grad_norm": 60.35252380371094,
+      "learning_rate": 2.7289999999999996e-07,
+      "loss": 0.4291,
+      "step": 181850
+    },
+    {
+      "epoch": 1.819,
+      "grad_norm": 28.022743225097656,
+      "learning_rate": 2.727e-07,
+      "loss": 0.3474,
+      "step": 181900
+    },
+    {
+      "epoch": 1.8195000000000001,
+      "grad_norm": 10.732136726379395,
+      "learning_rate": 2.725e-07,
+      "loss": 0.4097,
+      "step": 181950
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 85.77149200439453,
+      "learning_rate": 2.7229999999999997e-07,
+      "loss": 0.4671,
+      "step": 182000
+    },
+    {
+      "epoch": 1.8205,
+      "grad_norm": 69.69800567626953,
+      "learning_rate": 2.721e-07,
+      "loss": 0.4068,
+      "step": 182050
+    },
+    {
+      "epoch": 1.821,
+      "grad_norm": 67.32820129394531,
+      "learning_rate": 2.7189999999999994e-07,
+      "loss": 0.434,
+      "step": 182100
+    },
+    {
+      "epoch": 1.8215,
+      "grad_norm": 46.35198974609375,
+      "learning_rate": 2.717e-07,
+      "loss": 0.4207,
+      "step": 182150
+    },
+    {
+      "epoch": 1.822,
+      "grad_norm": 13.979206085205078,
+      "learning_rate": 2.715e-07,
+      "loss": 0.4041,
+      "step": 182200
+    },
+    {
+      "epoch": 1.8225,
+      "grad_norm": 3.441556453704834,
+      "learning_rate": 2.7129999999999995e-07,
+      "loss": 0.4985,
+      "step": 182250
+    },
+    {
+      "epoch": 1.823,
+      "grad_norm": 2.4120137691497803,
+      "learning_rate": 2.711e-07,
+      "loss": 0.5843,
+      "step": 182300
+    },
+    {
+      "epoch": 1.8235000000000001,
+      "grad_norm": 70.7459716796875,
+      "learning_rate": 2.7089999999999997e-07,
+      "loss": 0.3521,
+      "step": 182350
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 35.97295379638672,
+      "learning_rate": 2.707e-07,
+      "loss": 0.5621,
+      "step": 182400
+    },
+    {
+      "epoch": 1.8245,
+      "grad_norm": 89.31857299804688,
+      "learning_rate": 2.705e-07,
+      "loss": 0.3875,
+      "step": 182450
+    },
+    {
+      "epoch": 1.825,
+      "grad_norm": 91.55574035644531,
+      "learning_rate": 2.703e-07,
+      "loss": 0.4778,
+      "step": 182500
+    },
+    {
+      "epoch": 1.8255,
+      "grad_norm": 4.394865036010742,
+      "learning_rate": 2.701e-07,
+      "loss": 0.3469,
+      "step": 182550
+    },
+    {
+      "epoch": 1.826,
+      "grad_norm": 3.864041566848755,
+      "learning_rate": 2.6989999999999995e-07,
+      "loss": 0.3886,
+      "step": 182600
+    },
+    {
+      "epoch": 1.8265,
+      "grad_norm": 12.288315773010254,
+      "learning_rate": 2.697e-07,
+      "loss": 0.4518,
+      "step": 182650
+    },
+    {
+      "epoch": 1.827,
+      "grad_norm": 2.9489903450012207,
+      "learning_rate": 2.695e-07,
+      "loss": 0.3015,
+      "step": 182700
+    },
+    {
+      "epoch": 1.8275000000000001,
+      "grad_norm": 58.115970611572266,
+      "learning_rate": 2.6929999999999996e-07,
+      "loss": 0.3234,
+      "step": 182750
+    },
+    {
+      "epoch": 1.8279999999999998,
+      "grad_norm": 62.53365707397461,
+      "learning_rate": 2.691e-07,
+      "loss": 0.4949,
+      "step": 182800
+    },
+    {
+      "epoch": 1.8285,
+      "grad_norm": 38.055843353271484,
+      "learning_rate": 2.68904e-07,
+      "loss": 0.4362,
+      "step": 182850
+    },
+    {
+      "epoch": 1.829,
+      "grad_norm": 114.48504638671875,
+      "learning_rate": 2.6870399999999997e-07,
+      "loss": 0.4096,
+      "step": 182900
+    },
+    {
+      "epoch": 1.8295,
+      "grad_norm": 56.97817611694336,
+      "learning_rate": 2.68504e-07,
+      "loss": 0.4706,
+      "step": 182950
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 100.63391876220703,
+      "learning_rate": 2.68304e-07,
+      "loss": 0.436,
+      "step": 183000
+    },
+    {
+      "epoch": 1.8305,
+      "grad_norm": 0.5884892344474792,
+      "learning_rate": 2.68104e-07,
+      "loss": 0.5211,
+      "step": 183050
+    },
+    {
+      "epoch": 1.831,
+      "grad_norm": 39.73606491088867,
+      "learning_rate": 2.6790399999999996e-07,
+      "loss": 0.3914,
+      "step": 183100
+    },
+    {
+      "epoch": 1.8315000000000001,
+      "grad_norm": 3.421617269515991,
+      "learning_rate": 2.67704e-07,
+      "loss": 0.3778,
+      "step": 183150
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 61.44535827636719,
+      "learning_rate": 2.67504e-07,
+      "loss": 0.5635,
+      "step": 183200
+    },
+    {
+      "epoch": 1.8325,
+      "grad_norm": 80.6475601196289,
+      "learning_rate": 2.6730399999999997e-07,
+      "loss": 0.491,
+      "step": 183250
+    },
+    {
+      "epoch": 1.833,
+      "grad_norm": 5.449416160583496,
+      "learning_rate": 2.67104e-07,
+      "loss": 0.3328,
+      "step": 183300
+    },
+    {
+      "epoch": 1.8335,
+      "grad_norm": 92.49075317382812,
+      "learning_rate": 2.6690399999999994e-07,
+      "loss": 0.4623,
+      "step": 183350
+    },
+    {
+      "epoch": 1.834,
+      "grad_norm": 7.8489990234375,
+      "learning_rate": 2.66704e-07,
+      "loss": 0.4183,
+      "step": 183400
+    },
+    {
+      "epoch": 1.8345,
+      "grad_norm": 10.370362281799316,
+      "learning_rate": 2.66504e-07,
+      "loss": 0.4563,
+      "step": 183450
+    },
+    {
+      "epoch": 1.835,
+      "grad_norm": 104.78883361816406,
+      "learning_rate": 2.6630399999999995e-07,
+      "loss": 0.4643,
+      "step": 183500
+    },
+    {
+      "epoch": 1.8355000000000001,
+      "grad_norm": 1.3594282865524292,
+      "learning_rate": 2.66104e-07,
+      "loss": 0.4734,
+      "step": 183550
+    },
+    {
+      "epoch": 1.8359999999999999,
+      "grad_norm": 1.5864472389221191,
+      "learning_rate": 2.6590399999999997e-07,
+      "loss": 0.3522,
+      "step": 183600
+    },
+    {
+      "epoch": 1.8365,
+      "grad_norm": 62.5477409362793,
+      "learning_rate": 2.65704e-07,
+      "loss": 0.3794,
+      "step": 183650
+    },
+    {
+      "epoch": 1.837,
+      "grad_norm": 53.555747985839844,
+      "learning_rate": 2.65504e-07,
+      "loss": 0.4603,
+      "step": 183700
+    },
+    {
+      "epoch": 1.8375,
+      "grad_norm": 81.1266098022461,
+      "learning_rate": 2.65304e-07,
+      "loss": 0.564,
+      "step": 183750
+    },
+    {
+      "epoch": 1.838,
+      "grad_norm": 40.866695404052734,
+      "learning_rate": 2.65104e-07,
+      "loss": 0.4407,
+      "step": 183800
+    },
+    {
+      "epoch": 1.8385,
+      "grad_norm": 78.56951904296875,
+      "learning_rate": 2.6490399999999995e-07,
+      "loss": 0.435,
+      "step": 183850
+    },
+    {
+      "epoch": 1.839,
+      "grad_norm": 6.286223888397217,
+      "learning_rate": 2.64704e-07,
+      "loss": 0.3209,
+      "step": 183900
+    },
+    {
+      "epoch": 1.8395000000000001,
+      "grad_norm": 21.405439376831055,
+      "learning_rate": 2.64504e-07,
+      "loss": 0.352,
+      "step": 183950
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 1.1134364604949951,
+      "learning_rate": 2.6430399999999996e-07,
+      "loss": 0.4044,
+      "step": 184000
+    },
+    {
+      "epoch": 1.8405,
+      "grad_norm": 0.49174389243125916,
+      "learning_rate": 2.64104e-07,
+      "loss": 0.2879,
+      "step": 184050
+    },
+    {
+      "epoch": 1.841,
+      "grad_norm": 7.193144798278809,
+      "learning_rate": 2.6390400000000003e-07,
+      "loss": 0.5735,
+      "step": 184100
+    },
+    {
+      "epoch": 1.8415,
+      "grad_norm": 109.04924774169922,
+      "learning_rate": 2.6370399999999996e-07,
+      "loss": 0.4442,
+      "step": 184150
+    },
+    {
+      "epoch": 1.842,
+      "grad_norm": 29.094871520996094,
+      "learning_rate": 2.63504e-07,
+      "loss": 0.5491,
+      "step": 184200
+    },
+    {
+      "epoch": 1.8425,
+      "grad_norm": 50.769309997558594,
+      "learning_rate": 2.63304e-07,
+      "loss": 0.3682,
+      "step": 184250
+    },
+    {
+      "epoch": 1.843,
+      "grad_norm": 93.74052429199219,
+      "learning_rate": 2.6310399999999997e-07,
+      "loss": 0.612,
+      "step": 184300
+    },
+    {
+      "epoch": 1.8435000000000001,
+      "grad_norm": 90.94390106201172,
+      "learning_rate": 2.62904e-07,
+      "loss": 0.5965,
+      "step": 184350
+    },
+    {
+      "epoch": 1.8439999999999999,
+      "grad_norm": 66.96981811523438,
+      "learning_rate": 2.62704e-07,
+      "loss": 0.4703,
+      "step": 184400
+    },
+    {
+      "epoch": 1.8445,
+      "grad_norm": 99.0244369506836,
+      "learning_rate": 2.62504e-07,
+      "loss": 0.4547,
+      "step": 184450
+    },
+    {
+      "epoch": 1.845,
+      "grad_norm": 75.41879272460938,
+      "learning_rate": 2.6230399999999996e-07,
+      "loss": 0.3646,
+      "step": 184500
+    },
+    {
+      "epoch": 1.8455,
+      "grad_norm": 15.815811157226562,
+      "learning_rate": 2.62104e-07,
+      "loss": 0.432,
+      "step": 184550
+    },
+    {
+      "epoch": 1.846,
+      "grad_norm": 86.7995376586914,
+      "learning_rate": 2.6190400000000004e-07,
+      "loss": 0.5396,
+      "step": 184600
+    },
+    {
+      "epoch": 1.8465,
+      "grad_norm": 43.598876953125,
+      "learning_rate": 2.6170399999999997e-07,
+      "loss": 0.4494,
+      "step": 184650
+    },
+    {
+      "epoch": 1.847,
+      "grad_norm": 55.9370231628418,
+      "learning_rate": 2.61504e-07,
+      "loss": 0.4301,
+      "step": 184700
+    },
+    {
+      "epoch": 1.8475000000000001,
+      "grad_norm": 91.31571197509766,
+      "learning_rate": 2.61304e-07,
+      "loss": 0.4079,
+      "step": 184750
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 2.9025886058807373,
+      "learning_rate": 2.61104e-07,
+      "loss": 0.4143,
+      "step": 184800
+    },
+    {
+      "epoch": 1.8485,
+      "grad_norm": 76.20918273925781,
+      "learning_rate": 2.60904e-07,
+      "loss": 0.3667,
+      "step": 184850
+    },
+    {
+      "epoch": 1.849,
+      "grad_norm": 65.22622680664062,
+      "learning_rate": 2.60704e-07,
+      "loss": 0.3588,
+      "step": 184900
+    },
+    {
+      "epoch": 1.8495,
+      "grad_norm": 100.45918273925781,
+      "learning_rate": 2.60504e-07,
+      "loss": 0.4632,
+      "step": 184950
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.1237618923187256,
+      "learning_rate": 2.6030399999999997e-07,
+      "loss": 0.5477,
+      "step": 185000
+    },
+    {
+      "epoch": 1.8505,
+      "grad_norm": 64.92867279052734,
+      "learning_rate": 2.60104e-07,
+      "loss": 0.4146,
+      "step": 185050
+    },
+    {
+      "epoch": 1.851,
+      "grad_norm": 5.72701358795166,
+      "learning_rate": 2.59904e-07,
+      "loss": 0.3746,
+      "step": 185100
+    },
+    {
+      "epoch": 1.8515000000000001,
+      "grad_norm": 86.76318359375,
+      "learning_rate": 2.59704e-07,
+      "loss": 0.4661,
+      "step": 185150
+    },
+    {
+      "epoch": 1.8519999999999999,
+      "grad_norm": 68.64383697509766,
+      "learning_rate": 2.59504e-07,
+      "loss": 0.4119,
+      "step": 185200
+    },
+    {
+      "epoch": 1.8525,
+      "grad_norm": 10.763185501098633,
+      "learning_rate": 2.5930399999999995e-07,
+      "loss": 0.414,
+      "step": 185250
+    },
+    {
+      "epoch": 1.853,
+      "grad_norm": 1.534547209739685,
+      "learning_rate": 2.59104e-07,
+      "loss": 0.3791,
+      "step": 185300
+    },
+    {
+      "epoch": 1.8535,
+      "grad_norm": 6.461805820465088,
+      "learning_rate": 2.5890400000000003e-07,
+      "loss": 0.5048,
+      "step": 185350
+    },
+    {
+      "epoch": 1.854,
+      "grad_norm": 109.25067138671875,
+      "learning_rate": 2.5870399999999996e-07,
+      "loss": 0.5874,
+      "step": 185400
+    },
+    {
+      "epoch": 1.8545,
+      "grad_norm": 24.06308364868164,
+      "learning_rate": 2.58504e-07,
+      "loss": 0.3391,
+      "step": 185450
+    },
+    {
+      "epoch": 1.855,
+      "grad_norm": 0.7882880568504333,
+      "learning_rate": 2.58304e-07,
+      "loss": 0.4764,
+      "step": 185500
+    },
+    {
+      "epoch": 1.8555000000000001,
+      "grad_norm": 26.011205673217773,
+      "learning_rate": 2.5810399999999997e-07,
+      "loss": 0.4689,
+      "step": 185550
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 111.33613586425781,
+      "learning_rate": 2.57904e-07,
+      "loss": 0.4482,
+      "step": 185600
+    },
+    {
+      "epoch": 1.8565,
+      "grad_norm": 7.982728958129883,
+      "learning_rate": 2.57704e-07,
+      "loss": 0.4772,
+      "step": 185650
+    },
+    {
+      "epoch": 1.857,
+      "grad_norm": 68.1218032836914,
+      "learning_rate": 2.57504e-07,
+      "loss": 0.4142,
+      "step": 185700
+    },
+    {
+      "epoch": 1.8575,
+      "grad_norm": 22.570636749267578,
+      "learning_rate": 2.5730399999999996e-07,
+      "loss": 0.3669,
+      "step": 185750
+    },
+    {
+      "epoch": 1.858,
+      "grad_norm": 53.14181137084961,
+      "learning_rate": 2.57104e-07,
+      "loss": 0.3772,
+      "step": 185800
+    },
+    {
+      "epoch": 1.8585,
+      "grad_norm": 94.5253677368164,
+      "learning_rate": 2.5690400000000004e-07,
+      "loss": 0.5068,
+      "step": 185850
+    },
+    {
+      "epoch": 1.859,
+      "grad_norm": 57.95703887939453,
+      "learning_rate": 2.5670399999999997e-07,
+      "loss": 0.4562,
+      "step": 185900
+    },
+    {
+      "epoch": 1.8595000000000002,
+      "grad_norm": 125.20692443847656,
+      "learning_rate": 2.56504e-07,
+      "loss": 0.5206,
+      "step": 185950
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 76.74690246582031,
+      "learning_rate": 2.56304e-07,
+      "loss": 0.6174,
+      "step": 186000
+    },
+    {
+      "epoch": 1.8605,
+      "grad_norm": 36.07761764526367,
+      "learning_rate": 2.56104e-07,
+      "loss": 0.5131,
+      "step": 186050
+    },
+    {
+      "epoch": 1.861,
+      "grad_norm": 105.98657989501953,
+      "learning_rate": 2.55904e-07,
+      "loss": 0.4608,
+      "step": 186100
+    },
+    {
+      "epoch": 1.8615,
+      "grad_norm": 62.74296569824219,
+      "learning_rate": 2.55704e-07,
+      "loss": 0.3984,
+      "step": 186150
+    },
+    {
+      "epoch": 1.862,
+      "grad_norm": 7.824224948883057,
+      "learning_rate": 2.55504e-07,
+      "loss": 0.4741,
+      "step": 186200
+    },
+    {
+      "epoch": 1.8625,
+      "grad_norm": 0.2132396399974823,
+      "learning_rate": 2.5530399999999997e-07,
+      "loss": 0.4148,
+      "step": 186250
+    },
+    {
+      "epoch": 1.863,
+      "grad_norm": 63.773040771484375,
+      "learning_rate": 2.55104e-07,
+      "loss": 0.3911,
+      "step": 186300
+    },
+    {
+      "epoch": 1.8635000000000002,
+      "grad_norm": 35.12382888793945,
+      "learning_rate": 2.54904e-07,
+      "loss": 0.3641,
+      "step": 186350
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 13.937716484069824,
+      "learning_rate": 2.54704e-07,
+      "loss": 0.5185,
+      "step": 186400
+    },
+    {
+      "epoch": 1.8645,
+      "grad_norm": 139.38047790527344,
+      "learning_rate": 2.54504e-07,
+      "loss": 0.3617,
+      "step": 186450
+    },
+    {
+      "epoch": 1.865,
+      "grad_norm": 54.645015716552734,
+      "learning_rate": 2.5430399999999995e-07,
+      "loss": 0.4411,
+      "step": 186500
+    },
+    {
+      "epoch": 1.8655,
+      "grad_norm": 77.72416687011719,
+      "learning_rate": 2.54104e-07,
+      "loss": 0.4681,
+      "step": 186550
+    },
+    {
+      "epoch": 1.866,
+      "grad_norm": 75.3111801147461,
+      "learning_rate": 2.53904e-07,
+      "loss": 0.3701,
+      "step": 186600
+    },
+    {
+      "epoch": 1.8665,
+      "grad_norm": 90.5988540649414,
+      "learning_rate": 2.5370399999999995e-07,
+      "loss": 0.415,
+      "step": 186650
+    },
+    {
+      "epoch": 1.867,
+      "grad_norm": 7.769338130950928,
+      "learning_rate": 2.53504e-07,
+      "loss": 0.2835,
+      "step": 186700
+    },
+    {
+      "epoch": 1.8675000000000002,
+      "grad_norm": 38.08205032348633,
+      "learning_rate": 2.53304e-07,
+      "loss": 0.5307,
+      "step": 186750
+    },
+    {
+      "epoch": 1.8679999999999999,
+      "grad_norm": 17.543153762817383,
+      "learning_rate": 2.5310399999999996e-07,
+      "loss": 0.3548,
+      "step": 186800
+    },
+    {
+      "epoch": 1.8685,
+      "grad_norm": 10.238409996032715,
+      "learning_rate": 2.52904e-07,
+      "loss": 0.5478,
+      "step": 186850
+    },
+    {
+      "epoch": 1.869,
+      "grad_norm": 73.03883361816406,
+      "learning_rate": 2.52704e-07,
+      "loss": 0.4317,
+      "step": 186900
+    },
+    {
+      "epoch": 1.8695,
+      "grad_norm": 50.91599655151367,
+      "learning_rate": 2.5250399999999997e-07,
+      "loss": 0.4353,
+      "step": 186950
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.1760220527648926,
+      "learning_rate": 2.5230399999999995e-07,
+      "loss": 0.4632,
+      "step": 187000
+    },
+    {
+      "epoch": 1.8705,
+      "grad_norm": 79.37187194824219,
+      "learning_rate": 2.52104e-07,
+      "loss": 0.4926,
+      "step": 187050
+    },
+    {
+      "epoch": 1.871,
+      "grad_norm": 60.9011344909668,
+      "learning_rate": 2.5190400000000003e-07,
+      "loss": 0.5786,
+      "step": 187100
+    },
+    {
+      "epoch": 1.8715000000000002,
+      "grad_norm": 108.72712707519531,
+      "learning_rate": 2.5170399999999996e-07,
+      "loss": 0.3917,
+      "step": 187150
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 34.94348907470703,
+      "learning_rate": 2.51504e-07,
+      "loss": 0.4065,
+      "step": 187200
+    },
+    {
+      "epoch": 1.8725,
+      "grad_norm": 18.728527069091797,
+      "learning_rate": 2.5130400000000004e-07,
+      "loss": 0.523,
+      "step": 187250
+    },
+    {
+      "epoch": 1.873,
+      "grad_norm": 68.10762023925781,
+      "learning_rate": 2.5110399999999997e-07,
+      "loss": 0.372,
+      "step": 187300
+    },
+    {
+      "epoch": 1.8735,
+      "grad_norm": 24.434104919433594,
+      "learning_rate": 2.50904e-07,
+      "loss": 0.3331,
+      "step": 187350
+    },
+    {
+      "epoch": 1.874,
+      "grad_norm": 27.29010772705078,
+      "learning_rate": 2.50708e-07,
+      "loss": 0.5148,
+      "step": 187400
+    },
+    {
+      "epoch": 1.8745,
+      "grad_norm": 19.589698791503906,
+      "learning_rate": 2.50508e-07,
+      "loss": 0.4057,
+      "step": 187450
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 2.283752679824829,
+      "learning_rate": 2.5031199999999997e-07,
+      "loss": 0.3063,
+      "step": 187500
+    },
+    {
+      "epoch": 1.8755,
+      "grad_norm": 21.889986038208008,
+      "learning_rate": 2.50112e-07,
+      "loss": 0.3292,
+      "step": 187550
+    },
+    {
+      "epoch": 1.876,
+      "grad_norm": 2.8711414337158203,
+      "learning_rate": 2.49912e-07,
+      "loss": 0.4056,
+      "step": 187600
+    },
+    {
+      "epoch": 1.8765,
+      "grad_norm": 53.26054763793945,
+      "learning_rate": 2.49712e-07,
+      "loss": 0.4772,
+      "step": 187650
+    },
+    {
+      "epoch": 1.877,
+      "grad_norm": 10.286738395690918,
+      "learning_rate": 2.49512e-07,
+      "loss": 0.4867,
+      "step": 187700
+    },
+    {
+      "epoch": 1.8775,
+      "grad_norm": 93.31888580322266,
+      "learning_rate": 2.49312e-07,
+      "loss": 0.3756,
+      "step": 187750
+    },
+    {
+      "epoch": 1.8780000000000001,
+      "grad_norm": 86.73455810546875,
+      "learning_rate": 2.49112e-07,
+      "loss": 0.4222,
+      "step": 187800
+    },
+    {
+      "epoch": 1.8784999999999998,
+      "grad_norm": 7.344048023223877,
+      "learning_rate": 2.4891199999999997e-07,
+      "loss": 0.4104,
+      "step": 187850
+    },
+    {
+      "epoch": 1.879,
+      "grad_norm": 18.443588256835938,
+      "learning_rate": 2.4871199999999996e-07,
+      "loss": 0.5056,
+      "step": 187900
+    },
+    {
+      "epoch": 1.8795,
+      "grad_norm": 20.430662155151367,
+      "learning_rate": 2.48512e-07,
+      "loss": 0.7031,
+      "step": 187950
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 98.71015930175781,
+      "learning_rate": 2.48312e-07,
+      "loss": 0.5261,
+      "step": 188000
+    },
+    {
+      "epoch": 1.8805,
+      "grad_norm": 103.14667510986328,
+      "learning_rate": 2.4811199999999997e-07,
+      "loss": 0.4343,
+      "step": 188050
+    },
+    {
+      "epoch": 1.881,
+      "grad_norm": 82.66529083251953,
+      "learning_rate": 2.47912e-07,
+      "loss": 0.4226,
+      "step": 188100
+    },
+    {
+      "epoch": 1.8815,
+      "grad_norm": 11.182500839233398,
+      "learning_rate": 2.47712e-07,
+      "loss": 0.5237,
+      "step": 188150
+    },
+    {
+      "epoch": 1.8820000000000001,
+      "grad_norm": 74.19300842285156,
+      "learning_rate": 2.4751199999999997e-07,
+      "loss": 0.5641,
+      "step": 188200
+    },
+    {
+      "epoch": 1.8824999999999998,
+      "grad_norm": 36.28932571411133,
+      "learning_rate": 2.47312e-07,
+      "loss": 0.4777,
+      "step": 188250
+    },
+    {
+      "epoch": 1.883,
+      "grad_norm": 80.10577392578125,
+      "learning_rate": 2.47112e-07,
+      "loss": 0.5638,
+      "step": 188300
+    },
+    {
+      "epoch": 1.8835,
+      "grad_norm": 48.6932373046875,
+      "learning_rate": 2.46912e-07,
+      "loss": 0.3765,
+      "step": 188350
+    },
+    {
+      "epoch": 1.884,
+      "grad_norm": 2.2516491413116455,
+      "learning_rate": 2.4671199999999997e-07,
+      "loss": 0.5383,
+      "step": 188400
+    },
+    {
+      "epoch": 1.8845,
+      "grad_norm": 51.581634521484375,
+      "learning_rate": 2.46512e-07,
+      "loss": 0.5575,
+      "step": 188450
+    },
+    {
+      "epoch": 1.885,
+      "grad_norm": 29.83396339416504,
+      "learning_rate": 2.46312e-07,
+      "loss": 0.3791,
+      "step": 188500
+    },
+    {
+      "epoch": 1.8855,
+      "grad_norm": 30.861713409423828,
+      "learning_rate": 2.46112e-07,
+      "loss": 0.3886,
+      "step": 188550
+    },
+    {
+      "epoch": 1.8860000000000001,
+      "grad_norm": 0.5609220862388611,
+      "learning_rate": 2.4591199999999996e-07,
+      "loss": 0.4782,
+      "step": 188600
+    },
+    {
+      "epoch": 1.8864999999999998,
+      "grad_norm": 99.23149108886719,
+      "learning_rate": 2.45712e-07,
+      "loss": 0.4121,
+      "step": 188650
+    },
+    {
+      "epoch": 1.887,
+      "grad_norm": 75.1487808227539,
+      "learning_rate": 2.45512e-07,
+      "loss": 0.3858,
+      "step": 188700
+    },
+    {
+      "epoch": 1.8875,
+      "grad_norm": 52.555572509765625,
+      "learning_rate": 2.4531199999999997e-07,
+      "loss": 0.4526,
+      "step": 188750
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 3.6657705307006836,
+      "learning_rate": 2.45112e-07,
+      "loss": 0.3795,
+      "step": 188800
+    },
+    {
+      "epoch": 1.8885,
+      "grad_norm": 53.775360107421875,
+      "learning_rate": 2.44912e-07,
+      "loss": 0.4186,
+      "step": 188850
+    },
+    {
+      "epoch": 1.889,
+      "grad_norm": 82.37354278564453,
+      "learning_rate": 2.4471200000000003e-07,
+      "loss": 0.5195,
+      "step": 188900
+    },
+    {
+      "epoch": 1.8895,
+      "grad_norm": 71.7086181640625,
+      "learning_rate": 2.44512e-07,
+      "loss": 0.4569,
+      "step": 188950
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 76.88038635253906,
+      "learning_rate": 2.44312e-07,
+      "loss": 0.3721,
+      "step": 189000
+    },
+    {
+      "epoch": 1.8904999999999998,
+      "grad_norm": 16.016008377075195,
+      "learning_rate": 2.44112e-07,
+      "loss": 0.4164,
+      "step": 189050
+    },
+    {
+      "epoch": 1.891,
+      "grad_norm": 71.51254272460938,
+      "learning_rate": 2.4391199999999997e-07,
+      "loss": 0.4399,
+      "step": 189100
+    },
+    {
+      "epoch": 1.8915,
+      "grad_norm": 9.580039024353027,
+      "learning_rate": 2.43712e-07,
+      "loss": 0.3662,
+      "step": 189150
+    },
+    {
+      "epoch": 1.892,
+      "grad_norm": 75.79584503173828,
+      "learning_rate": 2.43512e-07,
+      "loss": 0.3707,
+      "step": 189200
+    },
+    {
+      "epoch": 1.8925,
+      "grad_norm": 86.73115539550781,
+      "learning_rate": 2.43312e-07,
+      "loss": 0.4684,
+      "step": 189250
+    },
+    {
+      "epoch": 1.893,
+      "grad_norm": 68.81775665283203,
+      "learning_rate": 2.4311199999999996e-07,
+      "loss": 0.3968,
+      "step": 189300
+    },
+    {
+      "epoch": 1.8935,
+      "grad_norm": 62.46356201171875,
+      "learning_rate": 2.42912e-07,
+      "loss": 0.3978,
+      "step": 189350
+    },
+    {
+      "epoch": 1.8940000000000001,
+      "grad_norm": 78.36255645751953,
+      "learning_rate": 2.42712e-07,
+      "loss": 0.4894,
+      "step": 189400
+    },
+    {
+      "epoch": 1.8944999999999999,
+      "grad_norm": 96.74028778076172,
+      "learning_rate": 2.4251199999999997e-07,
+      "loss": 0.3728,
+      "step": 189450
+    },
+    {
+      "epoch": 1.895,
+      "grad_norm": 92.6654052734375,
+      "learning_rate": 2.42312e-07,
+      "loss": 0.3868,
+      "step": 189500
+    },
+    {
+      "epoch": 1.8955,
+      "grad_norm": 38.866249084472656,
+      "learning_rate": 2.42112e-07,
+      "loss": 0.3451,
+      "step": 189550
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 63.626548767089844,
+      "learning_rate": 2.41912e-07,
+      "loss": 0.4468,
+      "step": 189600
+    },
+    {
+      "epoch": 1.8965,
+      "grad_norm": 65.61091613769531,
+      "learning_rate": 2.41712e-07,
+      "loss": 0.3838,
+      "step": 189650
+    },
+    {
+      "epoch": 1.897,
+      "grad_norm": 7.437851428985596,
+      "learning_rate": 2.41512e-07,
+      "loss": 0.3055,
+      "step": 189700
+    },
+    {
+      "epoch": 1.8975,
+      "grad_norm": 115.52892303466797,
+      "learning_rate": 2.41312e-07,
+      "loss": 0.3372,
+      "step": 189750
+    },
+    {
+      "epoch": 1.8980000000000001,
+      "grad_norm": 18.40287971496582,
+      "learning_rate": 2.4111199999999997e-07,
+      "loss": 0.3989,
+      "step": 189800
+    },
+    {
+      "epoch": 1.8984999999999999,
+      "grad_norm": 9.118658065795898,
+      "learning_rate": 2.4091199999999995e-07,
+      "loss": 0.4081,
+      "step": 189850
+    },
+    {
+      "epoch": 1.899,
+      "grad_norm": 18.847917556762695,
+      "learning_rate": 2.40712e-07,
+      "loss": 0.395,
+      "step": 189900
+    },
+    {
+      "epoch": 1.8995,
+      "grad_norm": 36.21388626098633,
+      "learning_rate": 2.40512e-07,
+      "loss": 0.3841,
+      "step": 189950
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 65.34298706054688,
+      "learning_rate": 2.4031199999999996e-07,
+      "loss": 0.4045,
+      "step": 190000
+    },
+    {
+      "epoch": 1.9005,
+      "grad_norm": 90.31636047363281,
+      "learning_rate": 2.40112e-07,
+      "loss": 0.3998,
+      "step": 190050
+    },
+    {
+      "epoch": 1.901,
+      "grad_norm": 80.00827026367188,
+      "learning_rate": 2.39912e-07,
+      "loss": 0.505,
+      "step": 190100
+    },
+    {
+      "epoch": 1.9015,
+      "grad_norm": 47.44596862792969,
+      "learning_rate": 2.39712e-07,
+      "loss": 0.3921,
+      "step": 190150
+    },
+    {
+      "epoch": 1.9020000000000001,
+      "grad_norm": 19.611328125,
+      "learning_rate": 2.39512e-07,
+      "loss": 0.3527,
+      "step": 190200
+    },
+    {
+      "epoch": 1.9024999999999999,
+      "grad_norm": 111.60746765136719,
+      "learning_rate": 2.39316e-07,
+      "loss": 0.4237,
+      "step": 190250
+    },
+    {
+      "epoch": 1.903,
+      "grad_norm": 0.5896288156509399,
+      "learning_rate": 2.39116e-07,
+      "loss": 0.6023,
+      "step": 190300
+    },
+    {
+      "epoch": 1.9035,
+      "grad_norm": 6.561922550201416,
+      "learning_rate": 2.3891599999999997e-07,
+      "loss": 0.4336,
+      "step": 190350
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 72.0843734741211,
+      "learning_rate": 2.38716e-07,
+      "loss": 0.4817,
+      "step": 190400
+    },
+    {
+      "epoch": 1.9045,
+      "grad_norm": 82.44991302490234,
+      "learning_rate": 2.38516e-07,
+      "loss": 0.5264,
+      "step": 190450
+    },
+    {
+      "epoch": 1.905,
+      "grad_norm": 65.08079528808594,
+      "learning_rate": 2.38316e-07,
+      "loss": 0.4905,
+      "step": 190500
+    },
+    {
+      "epoch": 1.9055,
+      "grad_norm": 52.838294982910156,
+      "learning_rate": 2.3811599999999999e-07,
+      "loss": 0.5025,
+      "step": 190550
+    },
+    {
+      "epoch": 1.9060000000000001,
+      "grad_norm": 16.74088478088379,
+      "learning_rate": 2.3791599999999997e-07,
+      "loss": 0.3235,
+      "step": 190600
+    },
+    {
+      "epoch": 1.9064999999999999,
+      "grad_norm": 37.48681640625,
+      "learning_rate": 2.37716e-07,
+      "loss": 0.4151,
+      "step": 190650
+    },
+    {
+      "epoch": 1.907,
+      "grad_norm": 84.55889892578125,
+      "learning_rate": 2.37516e-07,
+      "loss": 0.3849,
+      "step": 190700
+    },
+    {
+      "epoch": 1.9075,
+      "grad_norm": 13.987595558166504,
+      "learning_rate": 2.3731599999999998e-07,
+      "loss": 0.3796,
+      "step": 190750
+    },
+    {
+      "epoch": 1.908,
+      "grad_norm": 63.06429672241211,
+      "learning_rate": 2.37116e-07,
+      "loss": 0.3547,
+      "step": 190800
+    },
+    {
+      "epoch": 1.9085,
+      "grad_norm": 41.80989456176758,
+      "learning_rate": 2.3691599999999998e-07,
+      "loss": 0.5286,
+      "step": 190850
+    },
+    {
+      "epoch": 1.909,
+      "grad_norm": 83.75984954833984,
+      "learning_rate": 2.36716e-07,
+      "loss": 0.3391,
+      "step": 190900
+    },
+    {
+      "epoch": 1.9095,
+      "grad_norm": 51.583290100097656,
+      "learning_rate": 2.36516e-07,
+      "loss": 0.3597,
+      "step": 190950
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 65.20770263671875,
+      "learning_rate": 2.3631599999999998e-07,
+      "loss": 0.5186,
+      "step": 191000
+    },
+    {
+      "epoch": 1.9104999999999999,
+      "grad_norm": 123.32764434814453,
+      "learning_rate": 2.36116e-07,
+      "loss": 0.4346,
+      "step": 191050
+    },
+    {
+      "epoch": 1.911,
+      "grad_norm": 41.057682037353516,
+      "learning_rate": 2.3591599999999998e-07,
+      "loss": 0.4873,
+      "step": 191100
+    },
+    {
+      "epoch": 1.9115,
+      "grad_norm": 40.01293182373047,
+      "learning_rate": 2.35716e-07,
+      "loss": 0.4443,
+      "step": 191150
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 0.16099533438682556,
+      "learning_rate": 2.35516e-07,
+      "loss": 0.3952,
+      "step": 191200
+    },
+    {
+      "epoch": 1.9125,
+      "grad_norm": 48.290771484375,
+      "learning_rate": 2.35316e-07,
+      "loss": 0.3768,
+      "step": 191250
+    },
+    {
+      "epoch": 1.913,
+      "grad_norm": 0.869463324546814,
+      "learning_rate": 2.3511599999999997e-07,
+      "loss": 0.5064,
+      "step": 191300
+    },
+    {
+      "epoch": 1.9135,
+      "grad_norm": 174.40245056152344,
+      "learning_rate": 2.34916e-07,
+      "loss": 0.4593,
+      "step": 191350
+    },
+    {
+      "epoch": 1.9140000000000001,
+      "grad_norm": 64.9591293334961,
+      "learning_rate": 2.34716e-07,
+      "loss": 0.4874,
+      "step": 191400
+    },
+    {
+      "epoch": 1.9144999999999999,
+      "grad_norm": 11.59118366241455,
+      "learning_rate": 2.3451599999999998e-07,
+      "loss": 0.5005,
+      "step": 191450
+    },
+    {
+      "epoch": 1.915,
+      "grad_norm": 108.15178680419922,
+      "learning_rate": 2.34316e-07,
+      "loss": 0.5062,
+      "step": 191500
+    },
+    {
+      "epoch": 1.9155,
+      "grad_norm": 0.83865886926651,
+      "learning_rate": 2.3411599999999998e-07,
+      "loss": 0.3885,
+      "step": 191550
+    },
+    {
+      "epoch": 1.916,
+      "grad_norm": 85.95867919921875,
+      "learning_rate": 2.3391600000000001e-07,
+      "loss": 0.3974,
+      "step": 191600
+    },
+    {
+      "epoch": 1.9165,
+      "grad_norm": 31.9445858001709,
+      "learning_rate": 2.33716e-07,
+      "loss": 0.34,
+      "step": 191650
+    },
+    {
+      "epoch": 1.917,
+      "grad_norm": 2.4149677753448486,
+      "learning_rate": 2.3351599999999998e-07,
+      "loss": 0.6457,
+      "step": 191700
+    },
+    {
+      "epoch": 1.9175,
+      "grad_norm": 18.962900161743164,
+      "learning_rate": 2.33316e-07,
+      "loss": 0.5091,
+      "step": 191750
+    },
+    {
+      "epoch": 1.9180000000000001,
+      "grad_norm": 73.12571716308594,
+      "learning_rate": 2.3311599999999998e-07,
+      "loss": 0.4173,
+      "step": 191800
+    },
+    {
+      "epoch": 1.9184999999999999,
+      "grad_norm": 5.990804195404053,
+      "learning_rate": 2.32916e-07,
+      "loss": 0.4436,
+      "step": 191850
+    },
+    {
+      "epoch": 1.919,
+      "grad_norm": 70.40878295898438,
+      "learning_rate": 2.32716e-07,
+      "loss": 0.4218,
+      "step": 191900
+    },
+    {
+      "epoch": 1.9195,
+      "grad_norm": 24.912353515625,
+      "learning_rate": 2.32516e-07,
+      "loss": 0.5248,
+      "step": 191950
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.189905166625977,
+      "learning_rate": 2.3231599999999997e-07,
+      "loss": 0.4552,
+      "step": 192000
+    },
+    {
+      "epoch": 1.9205,
+      "grad_norm": 56.80104064941406,
+      "learning_rate": 2.3211599999999999e-07,
+      "loss": 0.5258,
+      "step": 192050
+    },
+    {
+      "epoch": 1.921,
+      "grad_norm": 69.89954376220703,
+      "learning_rate": 2.31916e-07,
+      "loss": 0.4205,
+      "step": 192100
+    },
+    {
+      "epoch": 1.9215,
+      "grad_norm": 72.8506851196289,
+      "learning_rate": 2.3171599999999998e-07,
+      "loss": 0.4343,
+      "step": 192150
+    },
+    {
+      "epoch": 1.9220000000000002,
+      "grad_norm": 77.9870376586914,
+      "learning_rate": 2.31516e-07,
+      "loss": 0.3266,
+      "step": 192200
+    },
+    {
+      "epoch": 1.9224999999999999,
+      "grad_norm": 128.5535430908203,
+      "learning_rate": 2.3131599999999998e-07,
+      "loss": 0.5274,
+      "step": 192250
+    },
+    {
+      "epoch": 1.923,
+      "grad_norm": 34.5822868347168,
+      "learning_rate": 2.31116e-07,
+      "loss": 0.4293,
+      "step": 192300
+    },
+    {
+      "epoch": 1.9235,
+      "grad_norm": 1.127679467201233,
+      "learning_rate": 2.30916e-07,
+      "loss": 0.5053,
+      "step": 192350
+    },
+    {
+      "epoch": 1.924,
+      "grad_norm": 122.0609130859375,
+      "learning_rate": 2.3071599999999999e-07,
+      "loss": 0.4999,
+      "step": 192400
+    },
+    {
+      "epoch": 1.9245,
+      "grad_norm": 4.369318962097168,
+      "learning_rate": 2.30516e-07,
+      "loss": 0.375,
+      "step": 192450
+    },
+    {
+      "epoch": 1.925,
+      "grad_norm": 16.14723014831543,
+      "learning_rate": 2.3031599999999998e-07,
+      "loss": 0.4943,
+      "step": 192500
+    },
+    {
+      "epoch": 1.9255,
+      "grad_norm": 9.635993957519531,
+      "learning_rate": 2.3011599999999997e-07,
+      "loss": 0.3829,
+      "step": 192550
+    },
+    {
+      "epoch": 1.9260000000000002,
+      "grad_norm": 37.19770050048828,
+      "learning_rate": 2.29916e-07,
+      "loss": 0.441,
+      "step": 192600
+    },
+    {
+      "epoch": 1.9264999999999999,
+      "grad_norm": 100.09040069580078,
+      "learning_rate": 2.29716e-07,
+      "loss": 0.4527,
+      "step": 192650
+    },
+    {
+      "epoch": 1.927,
+      "grad_norm": 73.9195785522461,
+      "learning_rate": 2.2951599999999998e-07,
+      "loss": 0.4453,
+      "step": 192700
+    },
+    {
+      "epoch": 1.9275,
+      "grad_norm": 5.231517314910889,
+      "learning_rate": 2.29316e-07,
+      "loss": 0.4418,
+      "step": 192750
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 86.26935577392578,
+      "learning_rate": 2.2911599999999997e-07,
+      "loss": 0.4282,
+      "step": 192800
+    },
+    {
+      "epoch": 1.9285,
+      "grad_norm": 20.850133895874023,
+      "learning_rate": 2.28916e-07,
+      "loss": 0.5655,
+      "step": 192850
+    },
+    {
+      "epoch": 1.929,
+      "grad_norm": 0.3520480692386627,
+      "learning_rate": 2.28716e-07,
+      "loss": 0.409,
+      "step": 192900
+    },
+    {
+      "epoch": 1.9295,
+      "grad_norm": 3.128180503845215,
+      "learning_rate": 2.2851599999999998e-07,
+      "loss": 0.4842,
+      "step": 192950
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 122.73941040039062,
+      "learning_rate": 2.28316e-07,
+      "loss": 0.4514,
+      "step": 193000
+    },
+    {
+      "epoch": 1.9304999999999999,
+      "grad_norm": 9.219717979431152,
+      "learning_rate": 2.28116e-07,
+      "loss": 0.3849,
+      "step": 193050
+    },
+    {
+      "epoch": 1.931,
+      "grad_norm": 12.25208854675293,
+      "learning_rate": 2.27916e-07,
+      "loss": 0.4238,
+      "step": 193100
+    },
+    {
+      "epoch": 1.9315,
+      "grad_norm": 92.9471435546875,
+      "learning_rate": 2.27716e-07,
+      "loss": 0.3785,
+      "step": 193150
+    },
+    {
+      "epoch": 1.932,
+      "grad_norm": 35.56284713745117,
+      "learning_rate": 2.2751599999999998e-07,
+      "loss": 0.4077,
+      "step": 193200
+    },
+    {
+      "epoch": 1.9325,
+      "grad_norm": 85.1362075805664,
+      "learning_rate": 2.2731599999999997e-07,
+      "loss": 0.4885,
+      "step": 193250
+    },
+    {
+      "epoch": 1.933,
+      "grad_norm": 17.76969337463379,
+      "learning_rate": 2.2711999999999998e-07,
+      "loss": 0.5319,
+      "step": 193300
+    },
+    {
+      "epoch": 1.9335,
+      "grad_norm": 77.77349853515625,
+      "learning_rate": 2.2692399999999997e-07,
+      "loss": 0.6296,
+      "step": 193350
+    },
+    {
+      "epoch": 1.9340000000000002,
+      "grad_norm": 1.1785967350006104,
+      "learning_rate": 2.2672399999999999e-07,
+      "loss": 0.4844,
+      "step": 193400
+    },
+    {
+      "epoch": 1.9344999999999999,
+      "grad_norm": 36.741268157958984,
+      "learning_rate": 2.26524e-07,
+      "loss": 0.3688,
+      "step": 193450
+    },
+    {
+      "epoch": 1.935,
+      "grad_norm": 83.8866195678711,
+      "learning_rate": 2.2632399999999998e-07,
+      "loss": 0.5385,
+      "step": 193500
+    },
+    {
+      "epoch": 1.9355,
+      "grad_norm": 67.07383728027344,
+      "learning_rate": 2.26124e-07,
+      "loss": 0.3809,
+      "step": 193550
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 61.0508918762207,
+      "learning_rate": 2.25924e-07,
+      "loss": 0.362,
+      "step": 193600
+    },
+    {
+      "epoch": 1.9365,
+      "grad_norm": 95.52999877929688,
+      "learning_rate": 2.25724e-07,
+      "loss": 0.4692,
+      "step": 193650
+    },
+    {
+      "epoch": 1.937,
+      "grad_norm": 2.942073345184326,
+      "learning_rate": 2.25524e-07,
+      "loss": 0.4602,
+      "step": 193700
+    },
+    {
+      "epoch": 1.9375,
+      "grad_norm": 80.84626770019531,
+      "learning_rate": 2.2532399999999999e-07,
+      "loss": 0.6496,
+      "step": 193750
+    },
+    {
+      "epoch": 1.938,
+      "grad_norm": 15.654928207397461,
+      "learning_rate": 2.2512399999999997e-07,
+      "loss": 0.3803,
+      "step": 193800
+    },
+    {
+      "epoch": 1.9385,
+      "grad_norm": 61.12825393676758,
+      "learning_rate": 2.24924e-07,
+      "loss": 0.4215,
+      "step": 193850
+    },
+    {
+      "epoch": 1.939,
+      "grad_norm": 64.433837890625,
+      "learning_rate": 2.24724e-07,
+      "loss": 0.394,
+      "step": 193900
+    },
+    {
+      "epoch": 1.9395,
+      "grad_norm": 7.063285827636719,
+      "learning_rate": 2.2452399999999998e-07,
+      "loss": 0.4053,
+      "step": 193950
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 67.11646270751953,
+      "learning_rate": 2.24324e-07,
+      "loss": 0.5138,
+      "step": 194000
+    },
+    {
+      "epoch": 1.9405000000000001,
+      "grad_norm": 28.027524948120117,
+      "learning_rate": 2.2412399999999998e-07,
+      "loss": 0.3441,
+      "step": 194050
+    },
+    {
+      "epoch": 1.9409999999999998,
+      "grad_norm": 124.71054077148438,
+      "learning_rate": 2.2392400000000001e-07,
+      "loss": 0.4387,
+      "step": 194100
+    },
+    {
+      "epoch": 1.9415,
+      "grad_norm": 53.5966911315918,
+      "learning_rate": 2.23724e-07,
+      "loss": 0.3617,
+      "step": 194150
+    },
+    {
+      "epoch": 1.942,
+      "grad_norm": 52.884151458740234,
+      "learning_rate": 2.2352399999999998e-07,
+      "loss": 0.5952,
+      "step": 194200
+    },
+    {
+      "epoch": 1.9425,
+      "grad_norm": 5.888815879821777,
+      "learning_rate": 2.23324e-07,
+      "loss": 0.4729,
+      "step": 194250
+    },
+    {
+      "epoch": 1.943,
+      "grad_norm": 50.12318420410156,
+      "learning_rate": 2.2312399999999998e-07,
+      "loss": 0.443,
+      "step": 194300
+    },
+    {
+      "epoch": 1.9435,
+      "grad_norm": 53.48051452636719,
+      "learning_rate": 2.22924e-07,
+      "loss": 0.4987,
+      "step": 194350
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 40.23417663574219,
+      "learning_rate": 2.22724e-07,
+      "loss": 0.477,
+      "step": 194400
+    },
+    {
+      "epoch": 1.9445000000000001,
+      "grad_norm": 0.8471763730049133,
+      "learning_rate": 2.22524e-07,
+      "loss": 0.5448,
+      "step": 194450
+    },
+    {
+      "epoch": 1.9449999999999998,
+      "grad_norm": 3.5994434356689453,
+      "learning_rate": 2.2232399999999997e-07,
+      "loss": 0.3093,
+      "step": 194500
+    },
+    {
+      "epoch": 1.9455,
+      "grad_norm": 88.96204376220703,
+      "learning_rate": 2.2212399999999998e-07,
+      "loss": 0.447,
+      "step": 194550
+    },
+    {
+      "epoch": 1.946,
+      "grad_norm": 3.4328107833862305,
+      "learning_rate": 2.21924e-07,
+      "loss": 0.4485,
+      "step": 194600
+    },
+    {
+      "epoch": 1.9465,
+      "grad_norm": 0.81305330991745,
+      "learning_rate": 2.21724e-07,
+      "loss": 0.3417,
+      "step": 194650
+    },
+    {
+      "epoch": 1.947,
+      "grad_norm": 14.793620109558105,
+      "learning_rate": 2.21524e-07,
+      "loss": 0.2682,
+      "step": 194700
+    },
+    {
+      "epoch": 1.9475,
+      "grad_norm": 71.23736572265625,
+      "learning_rate": 2.2132399999999998e-07,
+      "loss": 0.5289,
+      "step": 194750
+    },
+    {
+      "epoch": 1.948,
+      "grad_norm": 89.07301330566406,
+      "learning_rate": 2.21124e-07,
+      "loss": 0.4109,
+      "step": 194800
+    },
+    {
+      "epoch": 1.9485000000000001,
+      "grad_norm": 169.87252807617188,
+      "learning_rate": 2.20924e-07,
+      "loss": 0.5863,
+      "step": 194850
+    },
+    {
+      "epoch": 1.9489999999999998,
+      "grad_norm": 107.96073150634766,
+      "learning_rate": 2.2072399999999999e-07,
+      "loss": 0.416,
+      "step": 194900
+    },
+    {
+      "epoch": 1.9495,
+      "grad_norm": 10.294443130493164,
+      "learning_rate": 2.20524e-07,
+      "loss": 0.4665,
+      "step": 194950
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 62.89139175415039,
+      "learning_rate": 2.2032399999999998e-07,
+      "loss": 0.5773,
+      "step": 195000
+    },
+    {
+      "epoch": 1.9505,
+      "grad_norm": 146.08712768554688,
+      "learning_rate": 2.2012399999999997e-07,
+      "loss": 0.5096,
+      "step": 195050
+    },
+    {
+      "epoch": 1.951,
+      "grad_norm": 5.977169990539551,
+      "learning_rate": 2.19924e-07,
+      "loss": 0.4419,
+      "step": 195100
+    },
+    {
+      "epoch": 1.9515,
+      "grad_norm": 34.73382568359375,
+      "learning_rate": 2.19724e-07,
+      "loss": 0.4073,
+      "step": 195150
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 111.31939697265625,
+      "learning_rate": 2.1952399999999997e-07,
+      "loss": 0.4008,
+      "step": 195200
+    },
+    {
+      "epoch": 1.9525000000000001,
+      "grad_norm": 28.672367095947266,
+      "learning_rate": 2.1932399999999999e-07,
+      "loss": 0.3195,
+      "step": 195250
+    },
+    {
+      "epoch": 1.9529999999999998,
+      "grad_norm": 51.09148025512695,
+      "learning_rate": 2.19124e-07,
+      "loss": 0.3417,
+      "step": 195300
+    },
+    {
+      "epoch": 1.9535,
+      "grad_norm": 13.32286262512207,
+      "learning_rate": 2.18924e-07,
+      "loss": 0.539,
+      "step": 195350
+    },
+    {
+      "epoch": 1.954,
+      "grad_norm": 89.55329895019531,
+      "learning_rate": 2.18724e-07,
+      "loss": 0.3611,
+      "step": 195400
+    },
+    {
+      "epoch": 1.9545,
+      "grad_norm": 112.88272857666016,
+      "learning_rate": 2.1852399999999998e-07,
+      "loss": 0.496,
+      "step": 195450
+    },
+    {
+      "epoch": 1.955,
+      "grad_norm": 34.98467254638672,
+      "learning_rate": 2.18324e-07,
+      "loss": 0.3647,
+      "step": 195500
+    },
+    {
+      "epoch": 1.9555,
+      "grad_norm": 25.51811981201172,
+      "learning_rate": 2.18124e-07,
+      "loss": 0.2741,
+      "step": 195550
+    },
+    {
+      "epoch": 1.956,
+      "grad_norm": 35.12691879272461,
+      "learning_rate": 2.1792399999999999e-07,
+      "loss": 0.5114,
+      "step": 195600
+    },
+    {
+      "epoch": 1.9565000000000001,
+      "grad_norm": 92.44993591308594,
+      "learning_rate": 2.17724e-07,
+      "loss": 0.4415,
+      "step": 195650
+    },
+    {
+      "epoch": 1.9569999999999999,
+      "grad_norm": 58.697418212890625,
+      "learning_rate": 2.1752399999999998e-07,
+      "loss": 0.5425,
+      "step": 195700
+    },
+    {
+      "epoch": 1.9575,
+      "grad_norm": 70.51270294189453,
+      "learning_rate": 2.1732399999999997e-07,
+      "loss": 0.4504,
+      "step": 195750
+    },
+    {
+      "epoch": 1.958,
+      "grad_norm": 100.30704498291016,
+      "learning_rate": 2.17124e-07,
+      "loss": 0.5119,
+      "step": 195800
+    },
+    {
+      "epoch": 1.9585,
+      "grad_norm": 74.18003845214844,
+      "learning_rate": 2.16924e-07,
+      "loss": 0.3975,
+      "step": 195850
+    },
+    {
+      "epoch": 1.959,
+      "grad_norm": 67.8548583984375,
+      "learning_rate": 2.16724e-07,
+      "loss": 0.3993,
+      "step": 195900
+    },
+    {
+      "epoch": 1.9595,
+      "grad_norm": 25.4310359954834,
+      "learning_rate": 2.16524e-07,
+      "loss": 0.4111,
+      "step": 195950
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.258418321609497,
+      "learning_rate": 2.1632399999999997e-07,
+      "loss": 0.4154,
+      "step": 196000
+    },
+    {
+      "epoch": 1.9605000000000001,
+      "grad_norm": 20.04966163635254,
+      "learning_rate": 2.16124e-07,
+      "loss": 0.4066,
+      "step": 196050
+    },
+    {
+      "epoch": 1.9609999999999999,
+      "grad_norm": 45.66105651855469,
+      "learning_rate": 2.15924e-07,
+      "loss": 0.503,
+      "step": 196100
+    },
+    {
+      "epoch": 1.9615,
+      "grad_norm": 46.578895568847656,
+      "learning_rate": 2.1572399999999998e-07,
+      "loss": 0.4432,
+      "step": 196150
+    },
+    {
+      "epoch": 1.962,
+      "grad_norm": 85.224609375,
+      "learning_rate": 2.15524e-07,
+      "loss": 0.5397,
+      "step": 196200
+    },
+    {
+      "epoch": 1.9625,
+      "grad_norm": 2.3414793014526367,
+      "learning_rate": 2.1532399999999998e-07,
+      "loss": 0.3125,
+      "step": 196250
+    },
+    {
+      "epoch": 1.963,
+      "grad_norm": 4.292512893676758,
+      "learning_rate": 2.15124e-07,
+      "loss": 0.4941,
+      "step": 196300
+    },
+    {
+      "epoch": 1.9635,
+      "grad_norm": 67.12167358398438,
+      "learning_rate": 2.14924e-07,
+      "loss": 0.4186,
+      "step": 196350
+    },
+    {
+      "epoch": 1.964,
+      "grad_norm": 74.41678619384766,
+      "learning_rate": 2.1472399999999998e-07,
+      "loss": 0.4799,
+      "step": 196400
+    },
+    {
+      "epoch": 1.9645000000000001,
+      "grad_norm": 55.400428771972656,
+      "learning_rate": 2.14524e-07,
+      "loss": 0.3932,
+      "step": 196450
+    },
+    {
+      "epoch": 1.9649999999999999,
+      "grad_norm": 89.4749526977539,
+      "learning_rate": 2.1432399999999998e-07,
+      "loss": 0.409,
+      "step": 196500
+    },
+    {
+      "epoch": 1.9655,
+      "grad_norm": 41.83777618408203,
+      "learning_rate": 2.14124e-07,
+      "loss": 0.3961,
+      "step": 196550
+    },
+    {
+      "epoch": 1.966,
+      "grad_norm": 46.31528854370117,
+      "learning_rate": 2.13924e-07,
+      "loss": 0.5221,
+      "step": 196600
+    },
+    {
+      "epoch": 1.9665,
+      "grad_norm": 64.41300964355469,
+      "learning_rate": 2.13724e-07,
+      "loss": 0.3646,
+      "step": 196650
+    },
+    {
+      "epoch": 1.967,
+      "grad_norm": 73.85462951660156,
+      "learning_rate": 2.1352399999999997e-07,
+      "loss": 0.496,
+      "step": 196700
+    },
+    {
+      "epoch": 1.9675,
+      "grad_norm": 15.591062545776367,
+      "learning_rate": 2.13324e-07,
+      "loss": 0.447,
+      "step": 196750
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 48.92272186279297,
+      "learning_rate": 2.13124e-07,
+      "loss": 0.4293,
+      "step": 196800
+    },
+    {
+      "epoch": 1.9685000000000001,
+      "grad_norm": 1.0661996603012085,
+      "learning_rate": 2.1292399999999998e-07,
+      "loss": 0.4461,
+      "step": 196850
+    },
+    {
+      "epoch": 1.9689999999999999,
+      "grad_norm": 3.930109739303589,
+      "learning_rate": 2.12724e-07,
+      "loss": 0.5336,
+      "step": 196900
+    },
+    {
+      "epoch": 1.9695,
+      "grad_norm": 59.66569519042969,
+      "learning_rate": 2.1252399999999998e-07,
+      "loss": 0.4064,
+      "step": 196950
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 59.22267532348633,
+      "learning_rate": 2.1232400000000002e-07,
+      "loss": 0.4626,
+      "step": 197000
+    },
+    {
+      "epoch": 1.9705,
+      "grad_norm": 61.249202728271484,
+      "learning_rate": 2.12124e-07,
+      "loss": 0.3603,
+      "step": 197050
+    },
+    {
+      "epoch": 1.971,
+      "grad_norm": 8.615995407104492,
+      "learning_rate": 2.1192399999999999e-07,
+      "loss": 0.4088,
+      "step": 197100
+    },
+    {
+      "epoch": 1.9715,
+      "grad_norm": 74.69889831542969,
+      "learning_rate": 2.11724e-07,
+      "loss": 0.4601,
+      "step": 197150
+    },
+    {
+      "epoch": 1.972,
+      "grad_norm": 70.49463653564453,
+      "learning_rate": 2.1152399999999998e-07,
+      "loss": 0.4344,
+      "step": 197200
+    },
+    {
+      "epoch": 1.9725000000000001,
+      "grad_norm": 62.61591339111328,
+      "learning_rate": 2.11324e-07,
+      "loss": 0.5932,
+      "step": 197250
+    },
+    {
+      "epoch": 1.9729999999999999,
+      "grad_norm": 25.135278701782227,
+      "learning_rate": 2.11124e-07,
+      "loss": 0.4321,
+      "step": 197300
+    },
+    {
+      "epoch": 1.9735,
+      "grad_norm": 1.134199619293213,
+      "learning_rate": 2.10924e-07,
+      "loss": 0.2776,
+      "step": 197350
+    },
+    {
+      "epoch": 1.974,
+      "grad_norm": 5.215174198150635,
+      "learning_rate": 2.1072399999999998e-07,
+      "loss": 0.3703,
+      "step": 197400
+    },
+    {
+      "epoch": 1.9745,
+      "grad_norm": 73.48294830322266,
+      "learning_rate": 2.1052399999999999e-07,
+      "loss": 0.5626,
+      "step": 197450
+    },
+    {
+      "epoch": 1.975,
+      "grad_norm": 136.4009552001953,
+      "learning_rate": 2.10324e-07,
+      "loss": 0.3805,
+      "step": 197500
+    },
+    {
+      "epoch": 1.9755,
+      "grad_norm": 79.76919555664062,
+      "learning_rate": 2.1012399999999998e-07,
+      "loss": 0.4289,
+      "step": 197550
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 70.65571594238281,
+      "learning_rate": 2.09924e-07,
+      "loss": 0.3265,
+      "step": 197600
+    },
+    {
+      "epoch": 1.9765000000000001,
+      "grad_norm": 83.61393737792969,
+      "learning_rate": 2.0972399999999998e-07,
+      "loss": 0.3778,
+      "step": 197650
+    },
+    {
+      "epoch": 1.9769999999999999,
+      "grad_norm": 0.6025605797767639,
+      "learning_rate": 2.09524e-07,
+      "loss": 0.4016,
+      "step": 197700
+    },
+    {
+      "epoch": 1.9775,
+      "grad_norm": 3.0272040367126465,
+      "learning_rate": 2.0932799999999998e-07,
+      "loss": 0.4813,
+      "step": 197750
+    },
+    {
+      "epoch": 1.978,
+      "grad_norm": 60.68215560913086,
+      "learning_rate": 2.09128e-07,
+      "loss": 0.353,
+      "step": 197800
+    },
+    {
+      "epoch": 1.9785,
+      "grad_norm": 140.40936279296875,
+      "learning_rate": 2.08928e-07,
+      "loss": 0.4967,
+      "step": 197850
+    },
+    {
+      "epoch": 1.979,
+      "grad_norm": 81.03306579589844,
+      "learning_rate": 2.08728e-07,
+      "loss": 0.6082,
+      "step": 197900
+    },
+    {
+      "epoch": 1.9795,
+      "grad_norm": 4.013239860534668,
+      "learning_rate": 2.0852799999999997e-07,
+      "loss": 0.4588,
+      "step": 197950
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 95.69232940673828,
+      "learning_rate": 2.08328e-07,
+      "loss": 0.4396,
+      "step": 198000
+    },
+    {
+      "epoch": 1.9805000000000001,
+      "grad_norm": 91.35757446289062,
+      "learning_rate": 2.08128e-07,
+      "loss": 0.4297,
+      "step": 198050
+    },
+    {
+      "epoch": 1.9809999999999999,
+      "grad_norm": 61.077537536621094,
+      "learning_rate": 2.0792799999999998e-07,
+      "loss": 0.4121,
+      "step": 198100
+    },
+    {
+      "epoch": 1.9815,
+      "grad_norm": 91.7123031616211,
+      "learning_rate": 2.07728e-07,
+      "loss": 0.3933,
+      "step": 198150
+    },
+    {
+      "epoch": 1.982,
+      "grad_norm": 58.250064849853516,
+      "learning_rate": 2.0752799999999998e-07,
+      "loss": 0.5652,
+      "step": 198200
+    },
+    {
+      "epoch": 1.9825,
+      "grad_norm": 4.037406921386719,
+      "learning_rate": 2.0732800000000002e-07,
+      "loss": 0.4317,
+      "step": 198250
+    },
+    {
+      "epoch": 1.983,
+      "grad_norm": 14.497562408447266,
+      "learning_rate": 2.07128e-07,
+      "loss": 0.3503,
+      "step": 198300
+    },
+    {
+      "epoch": 1.9835,
+      "grad_norm": 15.300369262695312,
+      "learning_rate": 2.0692799999999999e-07,
+      "loss": 0.4021,
+      "step": 198350
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 11.134997367858887,
+      "learning_rate": 2.06728e-07,
+      "loss": 0.45,
+      "step": 198400
+    },
+    {
+      "epoch": 1.9845000000000002,
+      "grad_norm": 17.25223731994629,
+      "learning_rate": 2.0652799999999998e-07,
+      "loss": 0.302,
+      "step": 198450
+    },
+    {
+      "epoch": 1.9849999999999999,
+      "grad_norm": 50.793540954589844,
+      "learning_rate": 2.06328e-07,
+      "loss": 0.4223,
+      "step": 198500
+    },
+    {
+      "epoch": 1.9855,
+      "grad_norm": 1.6654607057571411,
+      "learning_rate": 2.06128e-07,
+      "loss": 0.4057,
+      "step": 198550
+    },
+    {
+      "epoch": 1.986,
+      "grad_norm": 15.111828804016113,
+      "learning_rate": 2.05928e-07,
+      "loss": 0.2903,
+      "step": 198600
+    },
+    {
+      "epoch": 1.9865,
+      "grad_norm": 60.819488525390625,
+      "learning_rate": 2.0572799999999997e-07,
+      "loss": 0.3924,
+      "step": 198650
+    },
+    {
+      "epoch": 1.987,
+      "grad_norm": 58.338111877441406,
+      "learning_rate": 2.0552799999999999e-07,
+      "loss": 0.4706,
+      "step": 198700
+    },
+    {
+      "epoch": 1.9875,
+      "grad_norm": 4.499456882476807,
+      "learning_rate": 2.05328e-07,
+      "loss": 0.4668,
+      "step": 198750
+    },
+    {
+      "epoch": 1.988,
+      "grad_norm": 26.65963363647461,
+      "learning_rate": 2.0512799999999998e-07,
+      "loss": 0.464,
+      "step": 198800
+    },
+    {
+      "epoch": 1.9885000000000002,
+      "grad_norm": 33.08699035644531,
+      "learning_rate": 2.04928e-07,
+      "loss": 0.4391,
+      "step": 198850
+    },
+    {
+      "epoch": 1.9889999999999999,
+      "grad_norm": 25.592788696289062,
+      "learning_rate": 2.0472799999999998e-07,
+      "loss": 0.3808,
+      "step": 198900
+    },
+    {
+      "epoch": 1.9895,
+      "grad_norm": 58.27476501464844,
+      "learning_rate": 2.04528e-07,
+      "loss": 0.5133,
+      "step": 198950
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 133.251220703125,
+      "learning_rate": 2.04328e-07,
+      "loss": 0.5155,
+      "step": 199000
+    },
+    {
+      "epoch": 1.9905,
+      "grad_norm": 115.85850524902344,
+      "learning_rate": 2.0412799999999999e-07,
+      "loss": 0.3483,
+      "step": 199050
+    },
+    {
+      "epoch": 1.991,
+      "grad_norm": 6.595485687255859,
+      "learning_rate": 2.03928e-07,
+      "loss": 0.449,
+      "step": 199100
+    },
+    {
+      "epoch": 1.9915,
+      "grad_norm": 19.415233612060547,
+      "learning_rate": 2.0372799999999998e-07,
+      "loss": 0.5,
+      "step": 199150
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 36.67066192626953,
+      "learning_rate": 2.0352799999999997e-07,
+      "loss": 0.3246,
+      "step": 199200
+    },
+    {
+      "epoch": 1.9925000000000002,
+      "grad_norm": 82.13230895996094,
+      "learning_rate": 2.03328e-07,
+      "loss": 0.442,
+      "step": 199250
+    },
+    {
+      "epoch": 1.9929999999999999,
+      "grad_norm": 29.641822814941406,
+      "learning_rate": 2.03128e-07,
+      "loss": 0.3297,
+      "step": 199300
+    },
+    {
+      "epoch": 1.9935,
+      "grad_norm": 78.03038787841797,
+      "learning_rate": 2.0292799999999998e-07,
+      "loss": 0.5838,
+      "step": 199350
+    },
+    {
+      "epoch": 1.994,
+      "grad_norm": 14.143548011779785,
+      "learning_rate": 2.02728e-07,
+      "loss": 0.5141,
+      "step": 199400
+    },
+    {
+      "epoch": 1.9945,
+      "grad_norm": 124.76140594482422,
+      "learning_rate": 2.02528e-07,
+      "loss": 0.5651,
+      "step": 199450
+    },
+    {
+      "epoch": 1.995,
+      "grad_norm": 27.242807388305664,
+      "learning_rate": 2.02328e-07,
+      "loss": 0.4477,
+      "step": 199500
+    },
+    {
+      "epoch": 1.9955,
+      "grad_norm": 55.86709976196289,
+      "learning_rate": 2.02128e-07,
+      "loss": 0.5458,
+      "step": 199550
+    },
+    {
+      "epoch": 1.996,
+      "grad_norm": 61.482025146484375,
+      "learning_rate": 2.0192799999999998e-07,
+      "loss": 0.42,
+      "step": 199600
+    },
+    {
+      "epoch": 1.9965000000000002,
+      "grad_norm": 21.021808624267578,
+      "learning_rate": 2.01728e-07,
+      "loss": 0.4657,
+      "step": 199650
+    },
+    {
+      "epoch": 1.9969999999999999,
+      "grad_norm": 65.79966735839844,
+      "learning_rate": 2.01528e-07,
+      "loss": 0.4816,
+      "step": 199700
+    },
+    {
+      "epoch": 1.9975,
+      "grad_norm": 74.81526947021484,
+      "learning_rate": 2.01328e-07,
+      "loss": 0.5426,
+      "step": 199750
+    },
+    {
+      "epoch": 1.998,
+      "grad_norm": 78.884765625,
+      "learning_rate": 2.01128e-07,
+      "loss": 0.4315,
+      "step": 199800
+    },
+    {
+      "epoch": 1.9985,
+      "grad_norm": 10.851948738098145,
+      "learning_rate": 2.0092799999999998e-07,
+      "loss": 0.4302,
+      "step": 199850
+    },
+    {
+      "epoch": 1.999,
+      "grad_norm": 3.022129774093628,
+      "learning_rate": 2.0072799999999997e-07,
+      "loss": 0.3712,
+      "step": 199900
+    },
+    {
+      "epoch": 1.9995,
+      "grad_norm": 24.550790786743164,
+      "learning_rate": 2.00528e-07,
+      "loss": 0.4664,
+      "step": 199950
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.006537437438965,
+      "learning_rate": 2.00328e-07,
+      "loss": 0.3405,
+      "step": 200000
+    },
+    {
+      "epoch": 2.0005,
+      "grad_norm": 3.087618827819824,
+      "learning_rate": 2.00128e-07,
+      "loss": 0.4074,
+      "step": 200050
+    },
+    {
+      "epoch": 2.001,
+      "grad_norm": 1.2099264860153198,
+      "learning_rate": 1.99928e-07,
+      "loss": 0.4809,
+      "step": 200100
+    },
+    {
+      "epoch": 2.0015,
+      "grad_norm": 3.2248778343200684,
+      "learning_rate": 1.9972799999999997e-07,
+      "loss": 0.3725,
+      "step": 200150
+    },
+    {
+      "epoch": 2.002,
+      "grad_norm": 39.72489929199219,
+      "learning_rate": 1.99528e-07,
+      "loss": 0.3364,
+      "step": 200200
+    },
+    {
+      "epoch": 2.0025,
+      "grad_norm": 79.86190032958984,
+      "learning_rate": 1.99328e-07,
+      "loss": 0.3408,
+      "step": 200250
+    },
+    {
+      "epoch": 2.003,
+      "grad_norm": 45.70669174194336,
+      "learning_rate": 1.9912799999999998e-07,
+      "loss": 0.4011,
+      "step": 200300
+    },
+    {
+      "epoch": 2.0035,
+      "grad_norm": 80.72676849365234,
+      "learning_rate": 1.98928e-07,
+      "loss": 0.5005,
+      "step": 200350
+    },
+    {
+      "epoch": 2.004,
+      "grad_norm": 54.08387756347656,
+      "learning_rate": 1.9872799999999998e-07,
+      "loss": 0.552,
+      "step": 200400
+    },
+    {
+      "epoch": 2.0045,
+      "grad_norm": 14.636670112609863,
+      "learning_rate": 1.98528e-07,
+      "loss": 0.3083,
+      "step": 200450
+    },
+    {
+      "epoch": 2.005,
+      "grad_norm": 65.19412994384766,
+      "learning_rate": 1.98328e-07,
+      "loss": 0.419,
+      "step": 200500
+    },
+    {
+      "epoch": 2.0055,
+      "grad_norm": 18.607166290283203,
+      "learning_rate": 1.9812799999999999e-07,
+      "loss": 0.3372,
+      "step": 200550
+    },
+    {
+      "epoch": 2.006,
+      "grad_norm": 5.587798595428467,
+      "learning_rate": 1.9792799999999997e-07,
+      "loss": 0.3789,
+      "step": 200600
+    },
+    {
+      "epoch": 2.0065,
+      "grad_norm": 9.782854080200195,
+      "learning_rate": 1.9772799999999998e-07,
+      "loss": 0.3832,
+      "step": 200650
+    },
+    {
+      "epoch": 2.007,
+      "grad_norm": 92.8857650756836,
+      "learning_rate": 1.97528e-07,
+      "loss": 0.3575,
+      "step": 200700
+    },
+    {
+      "epoch": 2.0075,
+      "grad_norm": 157.7519073486328,
+      "learning_rate": 1.97332e-07,
+      "loss": 0.5908,
+      "step": 200750
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 103.3365478515625,
+      "learning_rate": 1.97132e-07,
+      "loss": 0.3179,
+      "step": 200800
+    },
+    {
+      "epoch": 2.0085,
+      "grad_norm": 42.684906005859375,
+      "learning_rate": 1.9693199999999998e-07,
+      "loss": 0.3399,
+      "step": 200850
+    },
+    {
+      "epoch": 2.009,
+      "grad_norm": 2.2768945693969727,
+      "learning_rate": 1.96732e-07,
+      "loss": 0.3791,
+      "step": 200900
+    },
+    {
+      "epoch": 2.0095,
+      "grad_norm": 8.21125316619873,
+      "learning_rate": 1.96532e-07,
+      "loss": 0.4884,
+      "step": 200950
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.066143035888672,
+      "learning_rate": 1.96332e-07,
+      "loss": 0.5262,
+      "step": 201000
+    },
+    {
+      "epoch": 2.0105,
+      "grad_norm": 62.395328521728516,
+      "learning_rate": 1.96132e-07,
+      "loss": 0.3039,
+      "step": 201050
+    },
+    {
+      "epoch": 2.011,
+      "grad_norm": 1.7823199033737183,
+      "learning_rate": 1.9593199999999998e-07,
+      "loss": 0.5953,
+      "step": 201100
+    },
+    {
+      "epoch": 2.0115,
+      "grad_norm": 84.18726348876953,
+      "learning_rate": 1.9573199999999997e-07,
+      "loss": 0.4766,
+      "step": 201150
+    },
+    {
+      "epoch": 2.012,
+      "grad_norm": 119.35281372070312,
+      "learning_rate": 1.95532e-07,
+      "loss": 0.4257,
+      "step": 201200
+    },
+    {
+      "epoch": 2.0125,
+      "grad_norm": 24.274431228637695,
+      "learning_rate": 1.95332e-07,
+      "loss": 0.5276,
+      "step": 201250
+    },
+    {
+      "epoch": 2.013,
+      "grad_norm": 28.234947204589844,
+      "learning_rate": 1.95132e-07,
+      "loss": 0.4958,
+      "step": 201300
+    },
+    {
+      "epoch": 2.0135,
+      "grad_norm": 36.3571891784668,
+      "learning_rate": 1.94932e-07,
+      "loss": 0.3327,
+      "step": 201350
+    },
+    {
+      "epoch": 2.014,
+      "grad_norm": 70.78429412841797,
+      "learning_rate": 1.9473199999999997e-07,
+      "loss": 0.3493,
+      "step": 201400
+    },
+    {
+      "epoch": 2.0145,
+      "grad_norm": 67.98007202148438,
+      "learning_rate": 1.94532e-07,
+      "loss": 0.5261,
+      "step": 201450
+    },
+    {
+      "epoch": 2.015,
+      "grad_norm": 2.617002487182617,
+      "learning_rate": 1.94332e-07,
+      "loss": 0.3285,
+      "step": 201500
+    },
+    {
+      "epoch": 2.0155,
+      "grad_norm": 27.248516082763672,
+      "learning_rate": 1.9413199999999998e-07,
+      "loss": 0.4242,
+      "step": 201550
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 85.0505142211914,
+      "learning_rate": 1.93932e-07,
+      "loss": 0.3836,
+      "step": 201600
+    },
+    {
+      "epoch": 2.0165,
+      "grad_norm": 14.559182167053223,
+      "learning_rate": 1.9373199999999998e-07,
+      "loss": 0.31,
+      "step": 201650
+    },
+    {
+      "epoch": 2.017,
+      "grad_norm": 82.49097442626953,
+      "learning_rate": 1.93532e-07,
+      "loss": 0.3404,
+      "step": 201700
+    },
+    {
+      "epoch": 2.0175,
+      "grad_norm": 46.396629333496094,
+      "learning_rate": 1.93332e-07,
+      "loss": 0.3666,
+      "step": 201750
+    },
+    {
+      "epoch": 2.018,
+      "grad_norm": 44.75120162963867,
+      "learning_rate": 1.9313199999999999e-07,
+      "loss": 0.4974,
+      "step": 201800
+    },
+    {
+      "epoch": 2.0185,
+      "grad_norm": 6.472919464111328,
+      "learning_rate": 1.92932e-07,
+      "loss": 0.2974,
+      "step": 201850
+    },
+    {
+      "epoch": 2.019,
+      "grad_norm": 41.595184326171875,
+      "learning_rate": 1.9273199999999998e-07,
+      "loss": 0.3399,
+      "step": 201900
+    },
+    {
+      "epoch": 2.0195,
+      "grad_norm": 119.87677001953125,
+      "learning_rate": 1.92532e-07,
+      "loss": 0.3771,
+      "step": 201950
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 85.57170104980469,
+      "learning_rate": 1.92336e-07,
+      "loss": 0.4158,
+      "step": 202000
+    },
+    {
+      "epoch": 2.0205,
+      "grad_norm": 29.789093017578125,
+      "learning_rate": 1.92136e-07,
+      "loss": 0.4293,
+      "step": 202050
+    },
+    {
+      "epoch": 2.021,
+      "grad_norm": 83.55206298828125,
+      "learning_rate": 1.9193599999999998e-07,
+      "loss": 0.4268,
+      "step": 202100
+    },
+    {
+      "epoch": 2.0215,
+      "grad_norm": 11.1190824508667,
+      "learning_rate": 1.91736e-07,
+      "loss": 0.4764,
+      "step": 202150
+    },
+    {
+      "epoch": 2.022,
+      "grad_norm": 36.4733772277832,
+      "learning_rate": 1.91536e-07,
+      "loss": 0.4419,
+      "step": 202200
+    },
+    {
+      "epoch": 2.0225,
+      "grad_norm": 73.61580657958984,
+      "learning_rate": 1.91336e-07,
+      "loss": 0.4973,
+      "step": 202250
+    },
+    {
+      "epoch": 2.023,
+      "grad_norm": 46.41292190551758,
+      "learning_rate": 1.91136e-07,
+      "loss": 0.4368,
+      "step": 202300
+    },
+    {
+      "epoch": 2.0235,
+      "grad_norm": 85.7939224243164,
+      "learning_rate": 1.9093599999999998e-07,
+      "loss": 0.2774,
+      "step": 202350
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 5.079582214355469,
+      "learning_rate": 1.9073599999999997e-07,
+      "loss": 0.3701,
+      "step": 202400
+    },
+    {
+      "epoch": 2.0245,
+      "grad_norm": 16.522323608398438,
+      "learning_rate": 1.90536e-07,
+      "loss": 0.4191,
+      "step": 202450
+    },
+    {
+      "epoch": 2.025,
+      "grad_norm": 44.48244857788086,
+      "learning_rate": 1.90336e-07,
+      "loss": 0.4953,
+      "step": 202500
+    },
+    {
+      "epoch": 2.0255,
+      "grad_norm": 46.651466369628906,
+      "learning_rate": 1.90136e-07,
+      "loss": 0.372,
+      "step": 202550
+    },
+    {
+      "epoch": 2.026,
+      "grad_norm": 8.931777000427246,
+      "learning_rate": 1.89936e-07,
+      "loss": 0.4437,
+      "step": 202600
+    },
+    {
+      "epoch": 2.0265,
+      "grad_norm": 11.248937606811523,
+      "learning_rate": 1.8973599999999997e-07,
+      "loss": 0.4143,
+      "step": 202650
+    },
+    {
+      "epoch": 2.027,
+      "grad_norm": 65.09107208251953,
+      "learning_rate": 1.89536e-07,
+      "loss": 0.3685,
+      "step": 202700
+    },
+    {
+      "epoch": 2.0275,
+      "grad_norm": 15.731978416442871,
+      "learning_rate": 1.89336e-07,
+      "loss": 0.5076,
+      "step": 202750
+    },
+    {
+      "epoch": 2.028,
+      "grad_norm": 117.5165023803711,
+      "learning_rate": 1.8913599999999998e-07,
+      "loss": 0.4142,
+      "step": 202800
+    },
+    {
+      "epoch": 2.0285,
+      "grad_norm": 8.751471519470215,
+      "learning_rate": 1.88936e-07,
+      "loss": 0.3654,
+      "step": 202850
+    },
+    {
+      "epoch": 2.029,
+      "grad_norm": 85.83335876464844,
+      "learning_rate": 1.8873599999999998e-07,
+      "loss": 0.4755,
+      "step": 202900
+    },
+    {
+      "epoch": 2.0295,
+      "grad_norm": 11.942030906677246,
+      "learning_rate": 1.88536e-07,
+      "loss": 0.4555,
+      "step": 202950
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0720551013946533,
+      "learning_rate": 1.88336e-07,
+      "loss": 0.3455,
+      "step": 203000
+    },
+    {
+      "epoch": 2.0305,
+      "grad_norm": 27.060739517211914,
+      "learning_rate": 1.8813599999999998e-07,
+      "loss": 0.4064,
+      "step": 203050
+    },
+    {
+      "epoch": 2.031,
+      "grad_norm": 12.625927925109863,
+      "learning_rate": 1.87936e-07,
+      "loss": 0.4275,
+      "step": 203100
+    },
+    {
+      "epoch": 2.0315,
+      "grad_norm": 16.742694854736328,
+      "learning_rate": 1.8773599999999998e-07,
+      "loss": 0.4582,
+      "step": 203150
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 77.07002258300781,
+      "learning_rate": 1.87536e-07,
+      "loss": 0.3919,
+      "step": 203200
+    },
+    {
+      "epoch": 2.0325,
+      "grad_norm": 64.95145416259766,
+      "learning_rate": 1.87336e-07,
+      "loss": 0.4802,
+      "step": 203250
+    },
+    {
+      "epoch": 2.033,
+      "grad_norm": 124.8387680053711,
+      "learning_rate": 1.87136e-07,
+      "loss": 0.5289,
+      "step": 203300
+    },
+    {
+      "epoch": 2.0335,
+      "grad_norm": 38.79643630981445,
+      "learning_rate": 1.8693599999999997e-07,
+      "loss": 0.3996,
+      "step": 203350
+    },
+    {
+      "epoch": 2.034,
+      "grad_norm": 64.64935302734375,
+      "learning_rate": 1.86736e-07,
+      "loss": 0.3993,
+      "step": 203400
+    },
+    {
+      "epoch": 2.0345,
+      "grad_norm": 62.82065963745117,
+      "learning_rate": 1.86536e-07,
+      "loss": 0.5019,
+      "step": 203450
+    },
+    {
+      "epoch": 2.035,
+      "grad_norm": 1.8743690252304077,
+      "learning_rate": 1.8633599999999998e-07,
+      "loss": 0.3265,
+      "step": 203500
+    },
+    {
+      "epoch": 2.0355,
+      "grad_norm": 136.758056640625,
+      "learning_rate": 1.86136e-07,
+      "loss": 0.3709,
+      "step": 203550
+    },
+    {
+      "epoch": 2.036,
+      "grad_norm": 93.76640319824219,
+      "learning_rate": 1.8593599999999998e-07,
+      "loss": 0.4507,
+      "step": 203600
+    },
+    {
+      "epoch": 2.0365,
+      "grad_norm": 2.7972347736358643,
+      "learning_rate": 1.8573600000000002e-07,
+      "loss": 0.4523,
+      "step": 203650
+    },
+    {
+      "epoch": 2.037,
+      "grad_norm": 12.453752517700195,
+      "learning_rate": 1.85536e-07,
+      "loss": 0.2875,
+      "step": 203700
+    },
+    {
+      "epoch": 2.0375,
+      "grad_norm": 77.1436996459961,
+      "learning_rate": 1.8533599999999999e-07,
+      "loss": 0.3493,
+      "step": 203750
+    },
+    {
+      "epoch": 2.038,
+      "grad_norm": 5.663081645965576,
+      "learning_rate": 1.85136e-07,
+      "loss": 0.4532,
+      "step": 203800
+    },
+    {
+      "epoch": 2.0385,
+      "grad_norm": 74.7771224975586,
+      "learning_rate": 1.8493599999999998e-07,
+      "loss": 0.3986,
+      "step": 203850
+    },
+    {
+      "epoch": 2.039,
+      "grad_norm": 170.24021911621094,
+      "learning_rate": 1.84736e-07,
+      "loss": 0.5018,
+      "step": 203900
+    },
+    {
+      "epoch": 2.0395,
+      "grad_norm": 4.51198148727417,
+      "learning_rate": 1.84536e-07,
+      "loss": 0.3477,
+      "step": 203950
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.396271228790283,
+      "learning_rate": 1.84336e-07,
+      "loss": 0.4144,
+      "step": 204000
+    },
+    {
+      "epoch": 2.0405,
+      "grad_norm": 38.89625549316406,
+      "learning_rate": 1.8413599999999998e-07,
+      "loss": 0.4672,
+      "step": 204050
+    },
+    {
+      "epoch": 2.041,
+      "grad_norm": 63.771202087402344,
+      "learning_rate": 1.83936e-07,
+      "loss": 0.4361,
+      "step": 204100
+    },
+    {
+      "epoch": 2.0415,
+      "grad_norm": 85.22208404541016,
+      "learning_rate": 1.83736e-07,
+      "loss": 0.3421,
+      "step": 204150
+    },
+    {
+      "epoch": 2.042,
+      "grad_norm": 16.77423095703125,
+      "learning_rate": 1.8353599999999998e-07,
+      "loss": 0.4003,
+      "step": 204200
+    },
+    {
+      "epoch": 2.0425,
+      "grad_norm": 6.526354789733887,
+      "learning_rate": 1.83336e-07,
+      "loss": 0.4882,
+      "step": 204250
+    },
+    {
+      "epoch": 2.043,
+      "grad_norm": 21.03365135192871,
+      "learning_rate": 1.8313599999999998e-07,
+      "loss": 0.4461,
+      "step": 204300
+    },
+    {
+      "epoch": 2.0435,
+      "grad_norm": 99.76431274414062,
+      "learning_rate": 1.82936e-07,
+      "loss": 0.3903,
+      "step": 204350
+    },
+    {
+      "epoch": 2.044,
+      "grad_norm": 40.31965637207031,
+      "learning_rate": 1.82736e-07,
+      "loss": 0.4469,
+      "step": 204400
+    },
+    {
+      "epoch": 2.0445,
+      "grad_norm": 32.08421325683594,
+      "learning_rate": 1.82536e-07,
+      "loss": 0.285,
+      "step": 204450
+    },
+    {
+      "epoch": 2.045,
+      "grad_norm": 106.2055892944336,
+      "learning_rate": 1.82336e-07,
+      "loss": 0.4367,
+      "step": 204500
+    },
+    {
+      "epoch": 2.0455,
+      "grad_norm": 81.09613037109375,
+      "learning_rate": 1.8213599999999998e-07,
+      "loss": 0.5206,
+      "step": 204550
+    },
+    {
+      "epoch": 2.046,
+      "grad_norm": 74.99372863769531,
+      "learning_rate": 1.8193599999999997e-07,
+      "loss": 0.4927,
+      "step": 204600
+    },
+    {
+      "epoch": 2.0465,
+      "grad_norm": 1.6157112121582031,
+      "learning_rate": 1.81736e-07,
+      "loss": 0.463,
+      "step": 204650
+    },
+    {
+      "epoch": 2.047,
+      "grad_norm": 95.24854278564453,
+      "learning_rate": 1.81536e-07,
+      "loss": 0.401,
+      "step": 204700
+    },
+    {
+      "epoch": 2.0475,
+      "grad_norm": 36.316490173339844,
+      "learning_rate": 1.8133599999999998e-07,
+      "loss": 0.4379,
+      "step": 204750
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 134.88839721679688,
+      "learning_rate": 1.81136e-07,
+      "loss": 0.3477,
+      "step": 204800
+    },
+    {
+      "epoch": 2.0485,
+      "grad_norm": 0.8851105570793152,
+      "learning_rate": 1.80936e-07,
+      "loss": 0.4119,
+      "step": 204850
+    },
+    {
+      "epoch": 2.049,
+      "grad_norm": 0.5529405474662781,
+      "learning_rate": 1.80736e-07,
+      "loss": 0.3685,
+      "step": 204900
+    },
+    {
+      "epoch": 2.0495,
+      "grad_norm": 2.562713146209717,
+      "learning_rate": 1.80536e-07,
+      "loss": 0.4975,
+      "step": 204950
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 90.19808197021484,
+      "learning_rate": 1.8033599999999998e-07,
+      "loss": 0.5049,
+      "step": 205000
+    },
+    {
+      "epoch": 2.0505,
+      "grad_norm": 39.76176071166992,
+      "learning_rate": 1.80136e-07,
+      "loss": 0.4066,
+      "step": 205050
+    },
+    {
+      "epoch": 2.051,
+      "grad_norm": 41.6057014465332,
+      "learning_rate": 1.79936e-07,
+      "loss": 0.5315,
+      "step": 205100
+    },
+    {
+      "epoch": 2.0515,
+      "grad_norm": 2.7808868885040283,
+      "learning_rate": 1.79736e-07,
+      "loss": 0.3613,
+      "step": 205150
+    },
+    {
+      "epoch": 2.052,
+      "grad_norm": 82.41030883789062,
+      "learning_rate": 1.79536e-07,
+      "loss": 0.407,
+      "step": 205200
+    },
+    {
+      "epoch": 2.0525,
+      "grad_norm": 60.45346450805664,
+      "learning_rate": 1.7933599999999999e-07,
+      "loss": 0.529,
+      "step": 205250
+    },
+    {
+      "epoch": 2.053,
+      "grad_norm": 41.536399841308594,
+      "learning_rate": 1.7913599999999997e-07,
+      "loss": 0.5233,
+      "step": 205300
+    },
+    {
+      "epoch": 2.0535,
+      "grad_norm": 18.10399055480957,
+      "learning_rate": 1.78936e-07,
+      "loss": 0.3914,
+      "step": 205350
+    },
+    {
+      "epoch": 2.054,
+      "grad_norm": 2.1177046298980713,
+      "learning_rate": 1.78736e-07,
+      "loss": 0.4353,
+      "step": 205400
+    },
+    {
+      "epoch": 2.0545,
+      "grad_norm": 2.495558500289917,
+      "learning_rate": 1.78536e-07,
+      "loss": 0.3869,
+      "step": 205450
+    },
+    {
+      "epoch": 2.055,
+      "grad_norm": 30.129528045654297,
+      "learning_rate": 1.78336e-07,
+      "loss": 0.4063,
+      "step": 205500
+    },
+    {
+      "epoch": 2.0555,
+      "grad_norm": 85.46407318115234,
+      "learning_rate": 1.7813599999999997e-07,
+      "loss": 0.548,
+      "step": 205550
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 63.010189056396484,
+      "learning_rate": 1.77936e-07,
+      "loss": 0.4037,
+      "step": 205600
+    },
+    {
+      "epoch": 2.0565,
+      "grad_norm": 23.553878784179688,
+      "learning_rate": 1.77736e-07,
+      "loss": 0.4054,
+      "step": 205650
+    },
+    {
+      "epoch": 2.057,
+      "grad_norm": 61.06326675415039,
+      "learning_rate": 1.7753599999999998e-07,
+      "loss": 0.3369,
+      "step": 205700
+    },
+    {
+      "epoch": 2.0575,
+      "grad_norm": 72.79309844970703,
+      "learning_rate": 1.77336e-07,
+      "loss": 0.4785,
+      "step": 205750
+    },
+    {
+      "epoch": 2.058,
+      "grad_norm": 83.79118347167969,
+      "learning_rate": 1.7713599999999998e-07,
+      "loss": 0.4321,
+      "step": 205800
+    },
+    {
+      "epoch": 2.0585,
+      "grad_norm": 91.41462707519531,
+      "learning_rate": 1.76936e-07,
+      "loss": 0.3816,
+      "step": 205850
+    },
+    {
+      "epoch": 2.059,
+      "grad_norm": 0.9627030491828918,
+      "learning_rate": 1.76736e-07,
+      "loss": 0.3933,
+      "step": 205900
+    },
+    {
+      "epoch": 2.0595,
+      "grad_norm": 76.58192443847656,
+      "learning_rate": 1.76536e-07,
+      "loss": 0.4329,
+      "step": 205950
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 92.036376953125,
+      "learning_rate": 1.7633599999999997e-07,
+      "loss": 0.4328,
+      "step": 206000
+    },
+    {
+      "epoch": 2.0605,
+      "grad_norm": 16.09295654296875,
+      "learning_rate": 1.7613599999999998e-07,
+      "loss": 0.4057,
+      "step": 206050
+    },
+    {
+      "epoch": 2.061,
+      "grad_norm": 20.907642364501953,
+      "learning_rate": 1.75936e-07,
+      "loss": 0.3151,
+      "step": 206100
+    },
+    {
+      "epoch": 2.0615,
+      "grad_norm": 125.13389587402344,
+      "learning_rate": 1.75736e-07,
+      "loss": 0.5345,
+      "step": 206150
+    },
+    {
+      "epoch": 2.062,
+      "grad_norm": 10.434032440185547,
+      "learning_rate": 1.75536e-07,
+      "loss": 0.4701,
+      "step": 206200
+    },
+    {
+      "epoch": 2.0625,
+      "grad_norm": 88.18071746826172,
+      "learning_rate": 1.7533599999999998e-07,
+      "loss": 0.6219,
+      "step": 206250
+    },
+    {
+      "epoch": 2.063,
+      "grad_norm": 107.26880645751953,
+      "learning_rate": 1.75136e-07,
+      "loss": 0.4287,
+      "step": 206300
+    },
+    {
+      "epoch": 2.0635,
+      "grad_norm": 64.41120147705078,
+      "learning_rate": 1.74936e-07,
+      "loss": 0.467,
+      "step": 206350
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 71.87296295166016,
+      "learning_rate": 1.7473599999999998e-07,
+      "loss": 0.5102,
+      "step": 206400
+    },
+    {
+      "epoch": 2.0645,
+      "grad_norm": 4.165521621704102,
+      "learning_rate": 1.74536e-07,
+      "loss": 0.5447,
+      "step": 206450
+    },
+    {
+      "epoch": 2.065,
+      "grad_norm": 10.817039489746094,
+      "learning_rate": 1.7433599999999998e-07,
+      "loss": 0.415,
+      "step": 206500
+    },
+    {
+      "epoch": 2.0655,
+      "grad_norm": 132.01255798339844,
+      "learning_rate": 1.7413600000000002e-07,
+      "loss": 0.4636,
+      "step": 206550
+    },
+    {
+      "epoch": 2.066,
+      "grad_norm": 4.783523082733154,
+      "learning_rate": 1.73936e-07,
+      "loss": 0.3517,
+      "step": 206600
+    },
+    {
+      "epoch": 2.0665,
+      "grad_norm": 91.72089385986328,
+      "learning_rate": 1.73736e-07,
+      "loss": 0.3424,
+      "step": 206650
+    },
+    {
+      "epoch": 2.067,
+      "grad_norm": 1.3586903810501099,
+      "learning_rate": 1.73536e-07,
+      "loss": 0.3763,
+      "step": 206700
+    },
+    {
+      "epoch": 2.0675,
+      "grad_norm": 119.87760162353516,
+      "learning_rate": 1.7333599999999998e-07,
+      "loss": 0.4029,
+      "step": 206750
+    },
+    {
+      "epoch": 2.068,
+      "grad_norm": 74.30902099609375,
+      "learning_rate": 1.73136e-07,
+      "loss": 0.4613,
+      "step": 206800
+    },
+    {
+      "epoch": 2.0685000000000002,
+      "grad_norm": 88.31536865234375,
+      "learning_rate": 1.72936e-07,
+      "loss": 0.4587,
+      "step": 206850
+    },
+    {
+      "epoch": 2.069,
+      "grad_norm": 30.35822105407715,
+      "learning_rate": 1.72736e-07,
+      "loss": 0.3296,
+      "step": 206900
+    },
+    {
+      "epoch": 2.0695,
+      "grad_norm": 68.1887435913086,
+      "learning_rate": 1.7253599999999998e-07,
+      "loss": 0.3588,
+      "step": 206950
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.027542114257812,
+      "learning_rate": 1.72336e-07,
+      "loss": 0.435,
+      "step": 207000
+    },
+    {
+      "epoch": 2.0705,
+      "grad_norm": 1.4884693622589111,
+      "learning_rate": 1.72136e-07,
+      "loss": 0.3958,
+      "step": 207050
+    },
+    {
+      "epoch": 2.071,
+      "grad_norm": 76.33940887451172,
+      "learning_rate": 1.7193599999999999e-07,
+      "loss": 0.4019,
+      "step": 207100
+    },
+    {
+      "epoch": 2.0715,
+      "grad_norm": 74.91709899902344,
+      "learning_rate": 1.71736e-07,
+      "loss": 0.4541,
+      "step": 207150
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 91.8819808959961,
+      "learning_rate": 1.7153599999999998e-07,
+      "loss": 0.5439,
+      "step": 207200
+    },
+    {
+      "epoch": 2.0725,
+      "grad_norm": 6.123153209686279,
+      "learning_rate": 1.71336e-07,
+      "loss": 0.4287,
+      "step": 207250
+    },
+    {
+      "epoch": 2.073,
+      "grad_norm": 0.9759462475776672,
+      "learning_rate": 1.71136e-07,
+      "loss": 0.374,
+      "step": 207300
+    },
+    {
+      "epoch": 2.0735,
+      "grad_norm": 15.24844741821289,
+      "learning_rate": 1.70936e-07,
+      "loss": 0.3756,
+      "step": 207350
+    },
+    {
+      "epoch": 2.074,
+      "grad_norm": 6.204489707946777,
+      "learning_rate": 1.70736e-07,
+      "loss": 0.4405,
+      "step": 207400
+    },
+    {
+      "epoch": 2.0745,
+      "grad_norm": 69.0223388671875,
+      "learning_rate": 1.7053599999999999e-07,
+      "loss": 0.4419,
+      "step": 207450
+    },
+    {
+      "epoch": 2.075,
+      "grad_norm": 39.172096252441406,
+      "learning_rate": 1.7033599999999997e-07,
+      "loss": 0.3817,
+      "step": 207500
+    },
+    {
+      "epoch": 2.0755,
+      "grad_norm": 59.810447692871094,
+      "learning_rate": 1.70136e-07,
+      "loss": 0.4333,
+      "step": 207550
+    },
+    {
+      "epoch": 2.076,
+      "grad_norm": 43.26020050048828,
+      "learning_rate": 1.69936e-07,
+      "loss": 0.4731,
+      "step": 207600
+    },
+    {
+      "epoch": 2.0765,
+      "grad_norm": 81.62408447265625,
+      "learning_rate": 1.6973599999999998e-07,
+      "loss": 0.4608,
+      "step": 207650
+    },
+    {
+      "epoch": 2.077,
+      "grad_norm": 0.3745923638343811,
+      "learning_rate": 1.69536e-07,
+      "loss": 0.3853,
+      "step": 207700
+    },
+    {
+      "epoch": 2.0775,
+      "grad_norm": 90.62264251708984,
+      "learning_rate": 1.6933599999999998e-07,
+      "loss": 0.4131,
+      "step": 207750
+    },
+    {
+      "epoch": 2.078,
+      "grad_norm": 12.992121696472168,
+      "learning_rate": 1.69136e-07,
+      "loss": 0.4822,
+      "step": 207800
+    },
+    {
+      "epoch": 2.0785,
+      "grad_norm": 54.97868347167969,
+      "learning_rate": 1.68936e-07,
+      "loss": 0.4623,
+      "step": 207850
+    },
+    {
+      "epoch": 2.079,
+      "grad_norm": 35.057308197021484,
+      "learning_rate": 1.6873599999999998e-07,
+      "loss": 0.2986,
+      "step": 207900
+    },
+    {
+      "epoch": 2.0795,
+      "grad_norm": 98.75119018554688,
+      "learning_rate": 1.68536e-07,
+      "loss": 0.5695,
+      "step": 207950
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 57.579368591308594,
+      "learning_rate": 1.68336e-07,
+      "loss": 0.5152,
+      "step": 208000
+    },
+    {
+      "epoch": 2.0805,
+      "grad_norm": 141.9813232421875,
+      "learning_rate": 1.68136e-07,
+      "loss": 0.4165,
+      "step": 208050
+    },
+    {
+      "epoch": 2.081,
+      "grad_norm": 114.23064422607422,
+      "learning_rate": 1.67936e-07,
+      "loss": 0.3704,
+      "step": 208100
+    },
+    {
+      "epoch": 2.0815,
+      "grad_norm": 160.40625,
+      "learning_rate": 1.67736e-07,
+      "loss": 0.3956,
+      "step": 208150
+    },
+    {
+      "epoch": 2.082,
+      "grad_norm": 47.36935043334961,
+      "learning_rate": 1.6753599999999997e-07,
+      "loss": 0.3628,
+      "step": 208200
+    },
+    {
+      "epoch": 2.0825,
+      "grad_norm": 114.87825012207031,
+      "learning_rate": 1.67336e-07,
+      "loss": 0.4358,
+      "step": 208250
+    },
+    {
+      "epoch": 2.083,
+      "grad_norm": 112.84601593017578,
+      "learning_rate": 1.67136e-07,
+      "loss": 0.4642,
+      "step": 208300
+    },
+    {
+      "epoch": 2.0835,
+      "grad_norm": 9.166009902954102,
+      "learning_rate": 1.66936e-07,
+      "loss": 0.3521,
+      "step": 208350
+    },
+    {
+      "epoch": 2.084,
+      "grad_norm": 114.99955749511719,
+      "learning_rate": 1.66736e-07,
+      "loss": 0.4923,
+      "step": 208400
+    },
+    {
+      "epoch": 2.0845,
+      "grad_norm": 3.3287408351898193,
+      "learning_rate": 1.6653599999999998e-07,
+      "loss": 0.3185,
+      "step": 208450
+    },
+    {
+      "epoch": 2.085,
+      "grad_norm": 6.908358097076416,
+      "learning_rate": 1.6634e-07,
+      "loss": 0.5219,
+      "step": 208500
+    },
+    {
+      "epoch": 2.0855,
+      "grad_norm": 24.365503311157227,
+      "learning_rate": 1.6614e-07,
+      "loss": 0.5059,
+      "step": 208550
+    },
+    {
+      "epoch": 2.086,
+      "grad_norm": 7.259128093719482,
+      "learning_rate": 1.6594e-07,
+      "loss": 0.4561,
+      "step": 208600
+    },
+    {
+      "epoch": 2.0865,
+      "grad_norm": 107.42555236816406,
+      "learning_rate": 1.6574e-07,
+      "loss": 0.3289,
+      "step": 208650
+    },
+    {
+      "epoch": 2.087,
+      "grad_norm": 7.124502182006836,
+      "learning_rate": 1.6553999999999999e-07,
+      "loss": 0.3946,
+      "step": 208700
+    },
+    {
+      "epoch": 2.0875,
+      "grad_norm": 71.81135559082031,
+      "learning_rate": 1.6533999999999997e-07,
+      "loss": 0.3836,
+      "step": 208750
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 95.23866271972656,
+      "learning_rate": 1.6514e-07,
+      "loss": 0.3454,
+      "step": 208800
+    },
+    {
+      "epoch": 2.0885,
+      "grad_norm": 20.035490036010742,
+      "learning_rate": 1.6494e-07,
+      "loss": 0.4794,
+      "step": 208850
+    },
+    {
+      "epoch": 2.089,
+      "grad_norm": 75.28366088867188,
+      "learning_rate": 1.6473999999999998e-07,
+      "loss": 0.5533,
+      "step": 208900
+    },
+    {
+      "epoch": 2.0895,
+      "grad_norm": 72.16234588623047,
+      "learning_rate": 1.6454e-07,
+      "loss": 0.4135,
+      "step": 208950
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 68.55174255371094,
+      "learning_rate": 1.6434e-07,
+      "loss": 0.3546,
+      "step": 209000
+    },
+    {
+      "epoch": 2.0905,
+      "grad_norm": 84.08557891845703,
+      "learning_rate": 1.6414e-07,
+      "loss": 0.4366,
+      "step": 209050
+    },
+    {
+      "epoch": 2.091,
+      "grad_norm": 119.96469116210938,
+      "learning_rate": 1.6394e-07,
+      "loss": 0.3998,
+      "step": 209100
+    },
+    {
+      "epoch": 2.0915,
+      "grad_norm": 15.461919784545898,
+      "learning_rate": 1.6373999999999998e-07,
+      "loss": 0.4224,
+      "step": 209150
+    },
+    {
+      "epoch": 2.092,
+      "grad_norm": 95.27957916259766,
+      "learning_rate": 1.6354e-07,
+      "loss": 0.4491,
+      "step": 209200
+    },
+    {
+      "epoch": 2.0925,
+      "grad_norm": 9.547892570495605,
+      "learning_rate": 1.6334e-07,
+      "loss": 0.4787,
+      "step": 209250
+    },
+    {
+      "epoch": 2.093,
+      "grad_norm": 56.198638916015625,
+      "learning_rate": 1.6314e-07,
+      "loss": 0.4518,
+      "step": 209300
+    },
+    {
+      "epoch": 2.0935,
+      "grad_norm": 45.14836502075195,
+      "learning_rate": 1.6294e-07,
+      "loss": 0.3812,
+      "step": 209350
+    },
+    {
+      "epoch": 2.094,
+      "grad_norm": 47.17598342895508,
+      "learning_rate": 1.6274e-07,
+      "loss": 0.3314,
+      "step": 209400
+    },
+    {
+      "epoch": 2.0945,
+      "grad_norm": 3.217527389526367,
+      "learning_rate": 1.6253999999999997e-07,
+      "loss": 0.4014,
+      "step": 209450
+    },
+    {
+      "epoch": 2.095,
+      "grad_norm": 59.638771057128906,
+      "learning_rate": 1.6234e-07,
+      "loss": 0.3964,
+      "step": 209500
+    },
+    {
+      "epoch": 2.0955,
+      "grad_norm": 83.21932220458984,
+      "learning_rate": 1.6214e-07,
+      "loss": 0.5206,
+      "step": 209550
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 15.194162368774414,
+      "learning_rate": 1.6194e-07,
+      "loss": 0.4676,
+      "step": 209600
+    },
+    {
+      "epoch": 2.0965,
+      "grad_norm": 107.33583068847656,
+      "learning_rate": 1.6174e-07,
+      "loss": 0.4261,
+      "step": 209650
+    },
+    {
+      "epoch": 2.097,
+      "grad_norm": 13.499813079833984,
+      "learning_rate": 1.6153999999999998e-07,
+      "loss": 0.5161,
+      "step": 209700
+    },
+    {
+      "epoch": 2.0975,
+      "grad_norm": 82.69535064697266,
+      "learning_rate": 1.6134000000000001e-07,
+      "loss": 0.4069,
+      "step": 209750
+    },
+    {
+      "epoch": 2.098,
+      "grad_norm": 41.087249755859375,
+      "learning_rate": 1.6114e-07,
+      "loss": 0.3837,
+      "step": 209800
+    },
+    {
+      "epoch": 2.0985,
+      "grad_norm": 19.804964065551758,
+      "learning_rate": 1.6093999999999998e-07,
+      "loss": 0.427,
+      "step": 209850
+    },
+    {
+      "epoch": 2.099,
+      "grad_norm": 36.766902923583984,
+      "learning_rate": 1.6074e-07,
+      "loss": 0.3817,
+      "step": 209900
+    },
+    {
+      "epoch": 2.0995,
+      "grad_norm": 128.2410125732422,
+      "learning_rate": 1.6053999999999998e-07,
+      "loss": 0.4625,
+      "step": 209950
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.3461100459098816,
+      "learning_rate": 1.6034e-07,
+      "loss": 0.4514,
+      "step": 210000
+    },
+    {
+      "epoch": 2.1005,
+      "grad_norm": 39.63062286376953,
+      "learning_rate": 1.6014e-07,
+      "loss": 0.4665,
+      "step": 210050
+    },
+    {
+      "epoch": 2.101,
+      "grad_norm": 29.11553382873535,
+      "learning_rate": 1.5994e-07,
+      "loss": 0.3503,
+      "step": 210100
+    },
+    {
+      "epoch": 2.1015,
+      "grad_norm": 18.823801040649414,
+      "learning_rate": 1.5973999999999997e-07,
+      "loss": 0.3922,
+      "step": 210150
+    },
+    {
+      "epoch": 2.102,
+      "grad_norm": 8.857904434204102,
+      "learning_rate": 1.5953999999999998e-07,
+      "loss": 0.4219,
+      "step": 210200
+    },
+    {
+      "epoch": 2.1025,
+      "grad_norm": 1.5444262027740479,
+      "learning_rate": 1.5934e-07,
+      "loss": 0.3796,
+      "step": 210250
+    },
+    {
+      "epoch": 2.103,
+      "grad_norm": 17.656465530395508,
+      "learning_rate": 1.59144e-07,
+      "loss": 0.5756,
+      "step": 210300
+    },
+    {
+      "epoch": 2.1035,
+      "grad_norm": 70.85519409179688,
+      "learning_rate": 1.58944e-07,
+      "loss": 0.3324,
+      "step": 210350
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 139.22637939453125,
+      "learning_rate": 1.5874399999999998e-07,
+      "loss": 0.5752,
+      "step": 210400
+    },
+    {
+      "epoch": 2.1045,
+      "grad_norm": 56.294315338134766,
+      "learning_rate": 1.58544e-07,
+      "loss": 0.4574,
+      "step": 210450
+    },
+    {
+      "epoch": 2.105,
+      "grad_norm": 45.46672821044922,
+      "learning_rate": 1.58344e-07,
+      "loss": 0.3502,
+      "step": 210500
+    },
+    {
+      "epoch": 2.1055,
+      "grad_norm": 11.471989631652832,
+      "learning_rate": 1.58144e-07,
+      "loss": 0.3842,
+      "step": 210550
+    },
+    {
+      "epoch": 2.106,
+      "grad_norm": 23.504499435424805,
+      "learning_rate": 1.57944e-07,
+      "loss": 0.4154,
+      "step": 210600
+    },
+    {
+      "epoch": 2.1065,
+      "grad_norm": 17.586835861206055,
+      "learning_rate": 1.5774399999999999e-07,
+      "loss": 0.4408,
+      "step": 210650
+    },
+    {
+      "epoch": 2.107,
+      "grad_norm": 74.19413757324219,
+      "learning_rate": 1.5754399999999997e-07,
+      "loss": 0.5318,
+      "step": 210700
+    },
+    {
+      "epoch": 2.1075,
+      "grad_norm": 5.282220840454102,
+      "learning_rate": 1.57344e-07,
+      "loss": 0.5633,
+      "step": 210750
+    },
+    {
+      "epoch": 2.108,
+      "grad_norm": 27.199790954589844,
+      "learning_rate": 1.57144e-07,
+      "loss": 0.3807,
+      "step": 210800
+    },
+    {
+      "epoch": 2.1085,
+      "grad_norm": 9.653837203979492,
+      "learning_rate": 1.56944e-07,
+      "loss": 0.5071,
+      "step": 210850
+    },
+    {
+      "epoch": 2.109,
+      "grad_norm": 2.2887659072875977,
+      "learning_rate": 1.56744e-07,
+      "loss": 0.442,
+      "step": 210900
+    },
+    {
+      "epoch": 2.1095,
+      "grad_norm": 130.85214233398438,
+      "learning_rate": 1.5654399999999998e-07,
+      "loss": 0.5694,
+      "step": 210950
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 71.07162475585938,
+      "learning_rate": 1.5634400000000001e-07,
+      "loss": 0.3873,
+      "step": 211000
+    },
+    {
+      "epoch": 2.1105,
+      "grad_norm": 0.9926930665969849,
+      "learning_rate": 1.56144e-07,
+      "loss": 0.4219,
+      "step": 211050
+    },
+    {
+      "epoch": 2.111,
+      "grad_norm": 31.29315948486328,
+      "learning_rate": 1.5594399999999998e-07,
+      "loss": 0.406,
+      "step": 211100
+    },
+    {
+      "epoch": 2.1115,
+      "grad_norm": 4.280275821685791,
+      "learning_rate": 1.55744e-07,
+      "loss": 0.4414,
+      "step": 211150
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 22.806076049804688,
+      "learning_rate": 1.5554399999999998e-07,
+      "loss": 0.4081,
+      "step": 211200
+    },
+    {
+      "epoch": 2.1125,
+      "grad_norm": 48.29012680053711,
+      "learning_rate": 1.55344e-07,
+      "loss": 0.4448,
+      "step": 211250
+    },
+    {
+      "epoch": 2.113,
+      "grad_norm": 28.17279624938965,
+      "learning_rate": 1.55144e-07,
+      "loss": 0.3505,
+      "step": 211300
+    },
+    {
+      "epoch": 2.1135,
+      "grad_norm": 4.374676704406738,
+      "learning_rate": 1.54944e-07,
+      "loss": 0.374,
+      "step": 211350
+    },
+    {
+      "epoch": 2.114,
+      "grad_norm": 5.979548931121826,
+      "learning_rate": 1.5474399999999997e-07,
+      "loss": 0.304,
+      "step": 211400
+    },
+    {
+      "epoch": 2.1145,
+      "grad_norm": 5.3794426918029785,
+      "learning_rate": 1.5454399999999998e-07,
+      "loss": 0.4963,
+      "step": 211450
+    },
+    {
+      "epoch": 2.115,
+      "grad_norm": 71.64678955078125,
+      "learning_rate": 1.54344e-07,
+      "loss": 0.4826,
+      "step": 211500
+    },
+    {
+      "epoch": 2.1155,
+      "grad_norm": 8.600757598876953,
+      "learning_rate": 1.54144e-07,
+      "loss": 0.3473,
+      "step": 211550
+    },
+    {
+      "epoch": 2.116,
+      "grad_norm": 6.017736434936523,
+      "learning_rate": 1.53944e-07,
+      "loss": 0.4768,
+      "step": 211600
+    },
+    {
+      "epoch": 2.1165,
+      "grad_norm": 58.22939682006836,
+      "learning_rate": 1.5374399999999998e-07,
+      "loss": 0.4373,
+      "step": 211650
+    },
+    {
+      "epoch": 2.117,
+      "grad_norm": 68.43115234375,
+      "learning_rate": 1.5354400000000001e-07,
+      "loss": 0.4691,
+      "step": 211700
+    },
+    {
+      "epoch": 2.1175,
+      "grad_norm": 79.82152557373047,
+      "learning_rate": 1.53344e-07,
+      "loss": 0.4244,
+      "step": 211750
+    },
+    {
+      "epoch": 2.118,
+      "grad_norm": 31.79875946044922,
+      "learning_rate": 1.5314399999999998e-07,
+      "loss": 0.5947,
+      "step": 211800
+    },
+    {
+      "epoch": 2.1185,
+      "grad_norm": 62.042083740234375,
+      "learning_rate": 1.52944e-07,
+      "loss": 0.3467,
+      "step": 211850
+    },
+    {
+      "epoch": 2.1189999999999998,
+      "grad_norm": 2.5608060359954834,
+      "learning_rate": 1.5274399999999998e-07,
+      "loss": 0.4214,
+      "step": 211900
+    },
+    {
+      "epoch": 2.1195,
+      "grad_norm": 105.99209594726562,
+      "learning_rate": 1.5254400000000002e-07,
+      "loss": 0.5052,
+      "step": 211950
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 24.472665786743164,
+      "learning_rate": 1.52344e-07,
+      "loss": 0.3036,
+      "step": 212000
+    },
+    {
+      "epoch": 2.1205,
+      "grad_norm": 97.40585327148438,
+      "learning_rate": 1.52144e-07,
+      "loss": 0.4331,
+      "step": 212050
+    },
+    {
+      "epoch": 2.121,
+      "grad_norm": 68.8331069946289,
+      "learning_rate": 1.51944e-07,
+      "loss": 0.4896,
+      "step": 212100
+    },
+    {
+      "epoch": 2.1215,
+      "grad_norm": 0.17462262511253357,
+      "learning_rate": 1.5174399999999999e-07,
+      "loss": 0.2989,
+      "step": 212150
+    },
+    {
+      "epoch": 2.122,
+      "grad_norm": 0.09120786935091019,
+      "learning_rate": 1.51544e-07,
+      "loss": 0.49,
+      "step": 212200
+    },
+    {
+      "epoch": 2.1225,
+      "grad_norm": 41.664825439453125,
+      "learning_rate": 1.51344e-07,
+      "loss": 0.4313,
+      "step": 212250
+    },
+    {
+      "epoch": 2.123,
+      "grad_norm": 99.65406036376953,
+      "learning_rate": 1.51144e-07,
+      "loss": 0.3627,
+      "step": 212300
+    },
+    {
+      "epoch": 2.1235,
+      "grad_norm": 71.01948547363281,
+      "learning_rate": 1.5094399999999998e-07,
+      "loss": 0.496,
+      "step": 212350
+    },
+    {
+      "epoch": 2.124,
+      "grad_norm": 90.17384338378906,
+      "learning_rate": 1.50744e-07,
+      "loss": 0.3975,
+      "step": 212400
+    },
+    {
+      "epoch": 2.1245,
+      "grad_norm": 16.80384635925293,
+      "learning_rate": 1.50544e-07,
+      "loss": 0.3572,
+      "step": 212450
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 63.9257926940918,
+      "learning_rate": 1.5034399999999999e-07,
+      "loss": 0.5707,
+      "step": 212500
+    },
+    {
+      "epoch": 2.1255,
+      "grad_norm": 40.26178741455078,
+      "learning_rate": 1.50144e-07,
+      "loss": 0.4961,
+      "step": 212550
+    },
+    {
+      "epoch": 2.126,
+      "grad_norm": 75.82178497314453,
+      "learning_rate": 1.4994399999999998e-07,
+      "loss": 0.4571,
+      "step": 212600
+    },
+    {
+      "epoch": 2.1265,
+      "grad_norm": 10.5368070602417,
+      "learning_rate": 1.49744e-07,
+      "loss": 0.4173,
+      "step": 212650
+    },
+    {
+      "epoch": 2.127,
+      "grad_norm": 141.17538452148438,
+      "learning_rate": 1.49544e-07,
+      "loss": 0.4677,
+      "step": 212700
+    },
+    {
+      "epoch": 2.1275,
+      "grad_norm": 58.944175720214844,
+      "learning_rate": 1.49344e-07,
+      "loss": 0.3237,
+      "step": 212750
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 43.46791458129883,
+      "learning_rate": 1.49144e-07,
+      "loss": 0.5176,
+      "step": 212800
+    },
+    {
+      "epoch": 2.1285,
+      "grad_norm": 1.668687105178833,
+      "learning_rate": 1.48944e-07,
+      "loss": 0.4732,
+      "step": 212850
+    },
+    {
+      "epoch": 2.129,
+      "grad_norm": 69.23137664794922,
+      "learning_rate": 1.4874399999999997e-07,
+      "loss": 0.4604,
+      "step": 212900
+    },
+    {
+      "epoch": 2.1295,
+      "grad_norm": 39.40574645996094,
+      "learning_rate": 1.48544e-07,
+      "loss": 0.4294,
+      "step": 212950
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 48.64951705932617,
+      "learning_rate": 1.48344e-07,
+      "loss": 0.3854,
+      "step": 213000
+    },
+    {
+      "epoch": 2.1305,
+      "grad_norm": 7.141056537628174,
+      "learning_rate": 1.4814399999999998e-07,
+      "loss": 0.3979,
+      "step": 213050
+    },
+    {
+      "epoch": 2.1310000000000002,
+      "grad_norm": 61.7899284362793,
+      "learning_rate": 1.47944e-07,
+      "loss": 0.4668,
+      "step": 213100
+    },
+    {
+      "epoch": 2.1315,
+      "grad_norm": 43.82136154174805,
+      "learning_rate": 1.4774399999999998e-07,
+      "loss": 0.4552,
+      "step": 213150
+    },
+    {
+      "epoch": 2.132,
+      "grad_norm": 9.3849458694458,
+      "learning_rate": 1.4754400000000001e-07,
+      "loss": 0.4971,
+      "step": 213200
+    },
+    {
+      "epoch": 2.1325,
+      "grad_norm": 0.2815798819065094,
+      "learning_rate": 1.47344e-07,
+      "loss": 0.3141,
+      "step": 213250
+    },
+    {
+      "epoch": 2.133,
+      "grad_norm": 70.74395751953125,
+      "learning_rate": 1.4714399999999998e-07,
+      "loss": 0.4281,
+      "step": 213300
+    },
+    {
+      "epoch": 2.1335,
+      "grad_norm": 36.044673919677734,
+      "learning_rate": 1.46944e-07,
+      "loss": 0.3749,
+      "step": 213350
+    },
+    {
+      "epoch": 2.134,
+      "grad_norm": 22.537832260131836,
+      "learning_rate": 1.46744e-07,
+      "loss": 0.4091,
+      "step": 213400
+    },
+    {
+      "epoch": 2.1345,
+      "grad_norm": 64.7277603149414,
+      "learning_rate": 1.46544e-07,
+      "loss": 0.4648,
+      "step": 213450
+    },
+    {
+      "epoch": 2.135,
+      "grad_norm": 32.96176528930664,
+      "learning_rate": 1.46344e-07,
+      "loss": 0.5325,
+      "step": 213500
+    },
+    {
+      "epoch": 2.1355,
+      "grad_norm": 39.71519470214844,
+      "learning_rate": 1.46144e-07,
+      "loss": 0.4756,
+      "step": 213550
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 27.864206314086914,
+      "learning_rate": 1.4594399999999997e-07,
+      "loss": 0.4653,
+      "step": 213600
+    },
+    {
+      "epoch": 2.1365,
+      "grad_norm": 59.47966766357422,
+      "learning_rate": 1.45744e-07,
+      "loss": 0.4488,
+      "step": 213650
+    },
+    {
+      "epoch": 2.137,
+      "grad_norm": 30.075504302978516,
+      "learning_rate": 1.45544e-07,
+      "loss": 0.5499,
+      "step": 213700
+    },
+    {
+      "epoch": 2.1375,
+      "grad_norm": 0.5983642339706421,
+      "learning_rate": 1.4534399999999998e-07,
+      "loss": 0.5613,
+      "step": 213750
+    },
+    {
+      "epoch": 2.138,
+      "grad_norm": 5.750205039978027,
+      "learning_rate": 1.45144e-07,
+      "loss": 0.5196,
+      "step": 213800
+    },
+    {
+      "epoch": 2.1385,
+      "grad_norm": 45.518558502197266,
+      "learning_rate": 1.4494399999999998e-07,
+      "loss": 0.2925,
+      "step": 213850
+    },
+    {
+      "epoch": 2.1390000000000002,
+      "grad_norm": 144.67080688476562,
+      "learning_rate": 1.4474400000000002e-07,
+      "loss": 0.4629,
+      "step": 213900
+    },
+    {
+      "epoch": 2.1395,
+      "grad_norm": 82.38689422607422,
+      "learning_rate": 1.44544e-07,
+      "loss": 0.5278,
+      "step": 213950
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.598920226097107,
+      "learning_rate": 1.4434399999999999e-07,
+      "loss": 0.2595,
+      "step": 214000
+    },
+    {
+      "epoch": 2.1405,
+      "grad_norm": 22.938703536987305,
+      "learning_rate": 1.44144e-07,
+      "loss": 0.4807,
+      "step": 214050
+    },
+    {
+      "epoch": 2.141,
+      "grad_norm": 50.35197067260742,
+      "learning_rate": 1.4394399999999998e-07,
+      "loss": 0.3687,
+      "step": 214100
+    },
+    {
+      "epoch": 2.1415,
+      "grad_norm": 56.01286697387695,
+      "learning_rate": 1.43744e-07,
+      "loss": 0.4494,
+      "step": 214150
+    },
+    {
+      "epoch": 2.142,
+      "grad_norm": 117.62816619873047,
+      "learning_rate": 1.43544e-07,
+      "loss": 0.4037,
+      "step": 214200
+    },
+    {
+      "epoch": 2.1425,
+      "grad_norm": 26.653152465820312,
+      "learning_rate": 1.43344e-07,
+      "loss": 0.3627,
+      "step": 214250
+    },
+    {
+      "epoch": 2.143,
+      "grad_norm": 18.058246612548828,
+      "learning_rate": 1.4314399999999997e-07,
+      "loss": 0.6102,
+      "step": 214300
+    },
+    {
+      "epoch": 2.1435,
+      "grad_norm": 6.283153057098389,
+      "learning_rate": 1.4294399999999999e-07,
+      "loss": 0.318,
+      "step": 214350
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 50.743133544921875,
+      "learning_rate": 1.42748e-07,
+      "loss": 0.3315,
+      "step": 214400
+    },
+    {
+      "epoch": 2.1445,
+      "grad_norm": 7.809654712677002,
+      "learning_rate": 1.4254800000000001e-07,
+      "loss": 0.3765,
+      "step": 214450
+    },
+    {
+      "epoch": 2.145,
+      "grad_norm": 58.41513442993164,
+      "learning_rate": 1.42348e-07,
+      "loss": 0.3844,
+      "step": 214500
+    },
+    {
+      "epoch": 2.1455,
+      "grad_norm": 117.92552185058594,
+      "learning_rate": 1.4214799999999998e-07,
+      "loss": 0.4366,
+      "step": 214550
+    },
+    {
+      "epoch": 2.146,
+      "grad_norm": 38.15401077270508,
+      "learning_rate": 1.41948e-07,
+      "loss": 0.4262,
+      "step": 214600
+    },
+    {
+      "epoch": 2.1465,
+      "grad_norm": 99.7507553100586,
+      "learning_rate": 1.41748e-07,
+      "loss": 0.4156,
+      "step": 214650
+    },
+    {
+      "epoch": 2.147,
+      "grad_norm": 120.47976684570312,
+      "learning_rate": 1.41548e-07,
+      "loss": 0.3704,
+      "step": 214700
+    },
+    {
+      "epoch": 2.1475,
+      "grad_norm": 0.7170859575271606,
+      "learning_rate": 1.41348e-07,
+      "loss": 0.5463,
+      "step": 214750
+    },
+    {
+      "epoch": 2.148,
+      "grad_norm": 73.54954528808594,
+      "learning_rate": 1.41148e-07,
+      "loss": 0.4351,
+      "step": 214800
+    },
+    {
+      "epoch": 2.1485,
+      "grad_norm": 1.4426411390304565,
+      "learning_rate": 1.4094799999999997e-07,
+      "loss": 0.4003,
+      "step": 214850
+    },
+    {
+      "epoch": 2.149,
+      "grad_norm": 16.82160758972168,
+      "learning_rate": 1.40748e-07,
+      "loss": 0.334,
+      "step": 214900
+    },
+    {
+      "epoch": 2.1495,
+      "grad_norm": 69.73733520507812,
+      "learning_rate": 1.40548e-07,
+      "loss": 0.427,
+      "step": 214950
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 125.05155944824219,
+      "learning_rate": 1.40348e-07,
+      "loss": 0.3633,
+      "step": 215000
+    },
+    {
+      "epoch": 2.1505,
+      "grad_norm": 55.94770812988281,
+      "learning_rate": 1.40148e-07,
+      "loss": 0.5265,
+      "step": 215050
+    },
+    {
+      "epoch": 2.151,
+      "grad_norm": 4.276247978210449,
+      "learning_rate": 1.3994799999999998e-07,
+      "loss": 0.3524,
+      "step": 215100
+    },
+    {
+      "epoch": 2.1515,
+      "grad_norm": 67.32160949707031,
+      "learning_rate": 1.3974800000000001e-07,
+      "loss": 0.4974,
+      "step": 215150
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 34.65168762207031,
+      "learning_rate": 1.39548e-07,
+      "loss": 0.3331,
+      "step": 215200
+    },
+    {
+      "epoch": 2.1525,
+      "grad_norm": 91.49262237548828,
+      "learning_rate": 1.3934799999999998e-07,
+      "loss": 0.4829,
+      "step": 215250
+    },
+    {
+      "epoch": 2.153,
+      "grad_norm": 14.011157989501953,
+      "learning_rate": 1.39148e-07,
+      "loss": 0.4288,
+      "step": 215300
+    },
+    {
+      "epoch": 2.1535,
+      "grad_norm": 5.634559631347656,
+      "learning_rate": 1.3894799999999998e-07,
+      "loss": 0.3349,
+      "step": 215350
+    },
+    {
+      "epoch": 2.154,
+      "grad_norm": 73.73646545410156,
+      "learning_rate": 1.38748e-07,
+      "loss": 0.4687,
+      "step": 215400
+    },
+    {
+      "epoch": 2.1545,
+      "grad_norm": 2.7727954387664795,
+      "learning_rate": 1.38548e-07,
+      "loss": 0.3353,
+      "step": 215450
+    },
+    {
+      "epoch": 2.155,
+      "grad_norm": 73.1183090209961,
+      "learning_rate": 1.38348e-07,
+      "loss": 0.5878,
+      "step": 215500
+    },
+    {
+      "epoch": 2.1555,
+      "grad_norm": 88.92938232421875,
+      "learning_rate": 1.3814799999999997e-07,
+      "loss": 0.3574,
+      "step": 215550
+    },
+    {
+      "epoch": 2.156,
+      "grad_norm": 9.608054161071777,
+      "learning_rate": 1.3794799999999999e-07,
+      "loss": 0.472,
+      "step": 215600
+    },
+    {
+      "epoch": 2.1565,
+      "grad_norm": 34.318458557128906,
+      "learning_rate": 1.37748e-07,
+      "loss": 0.3066,
+      "step": 215650
+    },
+    {
+      "epoch": 2.157,
+      "grad_norm": 146.12513732910156,
+      "learning_rate": 1.37548e-07,
+      "loss": 0.3393,
+      "step": 215700
+    },
+    {
+      "epoch": 2.1575,
+      "grad_norm": 7.567634582519531,
+      "learning_rate": 1.37348e-07,
+      "loss": 0.3595,
+      "step": 215750
+    },
+    {
+      "epoch": 2.158,
+      "grad_norm": 14.484745979309082,
+      "learning_rate": 1.3714799999999998e-07,
+      "loss": 0.3492,
+      "step": 215800
+    },
+    {
+      "epoch": 2.1585,
+      "grad_norm": 29.185087203979492,
+      "learning_rate": 1.36948e-07,
+      "loss": 0.4477,
+      "step": 215850
+    },
+    {
+      "epoch": 2.159,
+      "grad_norm": 3.4553093910217285,
+      "learning_rate": 1.36748e-07,
+      "loss": 0.6068,
+      "step": 215900
+    },
+    {
+      "epoch": 2.1595,
+      "grad_norm": 14.842061996459961,
+      "learning_rate": 1.3654799999999999e-07,
+      "loss": 0.4224,
+      "step": 215950
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 94.66940307617188,
+      "learning_rate": 1.36348e-07,
+      "loss": 0.4909,
+      "step": 216000
+    },
+    {
+      "epoch": 2.1605,
+      "grad_norm": 12.455718994140625,
+      "learning_rate": 1.3614799999999998e-07,
+      "loss": 0.4814,
+      "step": 216050
+    },
+    {
+      "epoch": 2.161,
+      "grad_norm": 118.89146423339844,
+      "learning_rate": 1.35948e-07,
+      "loss": 0.5104,
+      "step": 216100
+    },
+    {
+      "epoch": 2.1615,
+      "grad_norm": 64.86026763916016,
+      "learning_rate": 1.35748e-07,
+      "loss": 0.4312,
+      "step": 216150
+    },
+    {
+      "epoch": 2.162,
+      "grad_norm": 8.784263610839844,
+      "learning_rate": 1.35548e-07,
+      "loss": 0.4906,
+      "step": 216200
+    },
+    {
+      "epoch": 2.1625,
+      "grad_norm": 90.33892822265625,
+      "learning_rate": 1.35348e-07,
+      "loss": 0.4301,
+      "step": 216250
+    },
+    {
+      "epoch": 2.163,
+      "grad_norm": 22.997182846069336,
+      "learning_rate": 1.35148e-07,
+      "loss": 0.2993,
+      "step": 216300
+    },
+    {
+      "epoch": 2.1635,
+      "grad_norm": 35.568603515625,
+      "learning_rate": 1.34948e-07,
+      "loss": 0.3766,
+      "step": 216350
+    },
+    {
+      "epoch": 2.164,
+      "grad_norm": 28.420759201049805,
+      "learning_rate": 1.34748e-07,
+      "loss": 0.5307,
+      "step": 216400
+    },
+    {
+      "epoch": 2.1645,
+      "grad_norm": 19.557161331176758,
+      "learning_rate": 1.34548e-07,
+      "loss": 0.5116,
+      "step": 216450
+    },
+    {
+      "epoch": 2.165,
+      "grad_norm": 23.2882022857666,
+      "learning_rate": 1.3434799999999998e-07,
+      "loss": 0.4442,
+      "step": 216500
+    },
+    {
+      "epoch": 2.1655,
+      "grad_norm": 22.49976348876953,
+      "learning_rate": 1.34148e-07,
+      "loss": 0.4319,
+      "step": 216550
+    },
+    {
+      "epoch": 2.166,
+      "grad_norm": 79.17230987548828,
+      "learning_rate": 1.33948e-07,
+      "loss": 0.4977,
+      "step": 216600
+    },
+    {
+      "epoch": 2.1665,
+      "grad_norm": 68.50972747802734,
+      "learning_rate": 1.33748e-07,
+      "loss": 0.3239,
+      "step": 216650
+    },
+    {
+      "epoch": 2.167,
+      "grad_norm": 57.847503662109375,
+      "learning_rate": 1.33548e-07,
+      "loss": 0.3964,
+      "step": 216700
+    },
+    {
+      "epoch": 2.1675,
+      "grad_norm": 90.66923522949219,
+      "learning_rate": 1.3334799999999998e-07,
+      "loss": 0.3266,
+      "step": 216750
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 10.624639511108398,
+      "learning_rate": 1.33148e-07,
+      "loss": 0.4769,
+      "step": 216800
+    },
+    {
+      "epoch": 2.1685,
+      "grad_norm": 70.89537048339844,
+      "learning_rate": 1.32948e-07,
+      "loss": 0.4799,
+      "step": 216850
+    },
+    {
+      "epoch": 2.169,
+      "grad_norm": 52.987213134765625,
+      "learning_rate": 1.32748e-07,
+      "loss": 0.4412,
+      "step": 216900
+    },
+    {
+      "epoch": 2.1695,
+      "grad_norm": 69.99027252197266,
+      "learning_rate": 1.32552e-07,
+      "loss": 0.4945,
+      "step": 216950
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 90.99432373046875,
+      "learning_rate": 1.32356e-07,
+      "loss": 0.4697,
+      "step": 217000
+    },
+    {
+      "epoch": 2.1705,
+      "grad_norm": 79.41846466064453,
+      "learning_rate": 1.3215599999999998e-07,
+      "loss": 0.4307,
+      "step": 217050
+    },
+    {
+      "epoch": 2.171,
+      "grad_norm": 86.42295837402344,
+      "learning_rate": 1.31956e-07,
+      "loss": 0.4369,
+      "step": 217100
+    },
+    {
+      "epoch": 2.1715,
+      "grad_norm": 64.85267639160156,
+      "learning_rate": 1.31756e-07,
+      "loss": 0.4694,
+      "step": 217150
+    },
+    {
+      "epoch": 2.172,
+      "grad_norm": 4.685230731964111,
+      "learning_rate": 1.31556e-07,
+      "loss": 0.3749,
+      "step": 217200
+    },
+    {
+      "epoch": 2.1725,
+      "grad_norm": 98.96492004394531,
+      "learning_rate": 1.31356e-07,
+      "loss": 0.4746,
+      "step": 217250
+    },
+    {
+      "epoch": 2.173,
+      "grad_norm": 58.136985778808594,
+      "learning_rate": 1.3115599999999999e-07,
+      "loss": 0.4622,
+      "step": 217300
+    },
+    {
+      "epoch": 2.1734999999999998,
+      "grad_norm": 33.713130950927734,
+      "learning_rate": 1.3095599999999997e-07,
+      "loss": 0.411,
+      "step": 217350
+    },
+    {
+      "epoch": 2.174,
+      "grad_norm": 5.95607328414917,
+      "learning_rate": 1.30756e-07,
+      "loss": 0.4671,
+      "step": 217400
+    },
+    {
+      "epoch": 2.1745,
+      "grad_norm": 44.79324722290039,
+      "learning_rate": 1.30556e-07,
+      "loss": 0.3069,
+      "step": 217450
+    },
+    {
+      "epoch": 2.175,
+      "grad_norm": 21.093141555786133,
+      "learning_rate": 1.30356e-07,
+      "loss": 0.3747,
+      "step": 217500
+    },
+    {
+      "epoch": 2.1755,
+      "grad_norm": 27.197893142700195,
+      "learning_rate": 1.30156e-07,
+      "loss": 0.4195,
+      "step": 217550
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 77.7616195678711,
+      "learning_rate": 1.2995599999999998e-07,
+      "loss": 0.3689,
+      "step": 217600
+    },
+    {
+      "epoch": 2.1765,
+      "grad_norm": 0.3446306586265564,
+      "learning_rate": 1.2975600000000001e-07,
+      "loss": 0.4223,
+      "step": 217650
+    },
+    {
+      "epoch": 2.177,
+      "grad_norm": 95.07039642333984,
+      "learning_rate": 1.29556e-07,
+      "loss": 0.3375,
+      "step": 217700
+    },
+    {
+      "epoch": 2.1775,
+      "grad_norm": 80.28645324707031,
+      "learning_rate": 1.2935599999999998e-07,
+      "loss": 0.4666,
+      "step": 217750
+    },
+    {
+      "epoch": 2.178,
+      "grad_norm": 57.870567321777344,
+      "learning_rate": 1.29156e-07,
+      "loss": 0.5303,
+      "step": 217800
+    },
+    {
+      "epoch": 2.1785,
+      "grad_norm": 6.648061752319336,
+      "learning_rate": 1.2895599999999998e-07,
+      "loss": 0.3876,
+      "step": 217850
+    },
+    {
+      "epoch": 2.179,
+      "grad_norm": 80.23861694335938,
+      "learning_rate": 1.28756e-07,
+      "loss": 0.3716,
+      "step": 217900
+    },
+    {
+      "epoch": 2.1795,
+      "grad_norm": 44.08333206176758,
+      "learning_rate": 1.28556e-07,
+      "loss": 0.4331,
+      "step": 217950
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 54.142154693603516,
+      "learning_rate": 1.28356e-07,
+      "loss": 0.4002,
+      "step": 218000
+    },
+    {
+      "epoch": 2.1805,
+      "grad_norm": 13.98410701751709,
+      "learning_rate": 1.28156e-07,
+      "loss": 0.4893,
+      "step": 218050
+    },
+    {
+      "epoch": 2.181,
+      "grad_norm": 67.61353302001953,
+      "learning_rate": 1.2795599999999998e-07,
+      "loss": 0.4332,
+      "step": 218100
+    },
+    {
+      "epoch": 2.1814999999999998,
+      "grad_norm": 3.3425474166870117,
+      "learning_rate": 1.27756e-07,
+      "loss": 0.4579,
+      "step": 218150
+    },
+    {
+      "epoch": 2.182,
+      "grad_norm": 25.27934455871582,
+      "learning_rate": 1.27556e-07,
+      "loss": 0.4718,
+      "step": 218200
+    },
+    {
+      "epoch": 2.1825,
+      "grad_norm": 74.17402648925781,
+      "learning_rate": 1.27356e-07,
+      "loss": 0.3044,
+      "step": 218250
+    },
+    {
+      "epoch": 2.183,
+      "grad_norm": 12.15510368347168,
+      "learning_rate": 1.2715599999999998e-07,
+      "loss": 0.3911,
+      "step": 218300
+    },
+    {
+      "epoch": 2.1835,
+      "grad_norm": 8.140380859375,
+      "learning_rate": 1.2695600000000002e-07,
+      "loss": 0.376,
+      "step": 218350
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 102.70777130126953,
+      "learning_rate": 1.26756e-07,
+      "loss": 0.3637,
+      "step": 218400
+    },
+    {
+      "epoch": 2.1845,
+      "grad_norm": 19.61964225769043,
+      "learning_rate": 1.2655599999999999e-07,
+      "loss": 0.5274,
+      "step": 218450
+    },
+    {
+      "epoch": 2.185,
+      "grad_norm": 14.926944732666016,
+      "learning_rate": 1.26356e-07,
+      "loss": 0.5191,
+      "step": 218500
+    },
+    {
+      "epoch": 2.1855,
+      "grad_norm": 37.19790267944336,
+      "learning_rate": 1.2615599999999998e-07,
+      "loss": 0.4175,
+      "step": 218550
+    },
+    {
+      "epoch": 2.186,
+      "grad_norm": 107.50064849853516,
+      "learning_rate": 1.2595600000000002e-07,
+      "loss": 0.3893,
+      "step": 218600
+    },
+    {
+      "epoch": 2.1865,
+      "grad_norm": 125.32817077636719,
+      "learning_rate": 1.25756e-07,
+      "loss": 0.442,
+      "step": 218650
+    },
+    {
+      "epoch": 2.187,
+      "grad_norm": 5.219124794006348,
+      "learning_rate": 1.25556e-07,
+      "loss": 0.4723,
+      "step": 218700
+    },
+    {
+      "epoch": 2.1875,
+      "grad_norm": 69.8603286743164,
+      "learning_rate": 1.25356e-07,
+      "loss": 0.3469,
+      "step": 218750
+    },
+    {
+      "epoch": 2.188,
+      "grad_norm": 61.861080169677734,
+      "learning_rate": 1.2515599999999999e-07,
+      "loss": 0.4402,
+      "step": 218800
+    },
+    {
+      "epoch": 2.1885,
+      "grad_norm": 27.252426147460938,
+      "learning_rate": 1.24956e-07,
+      "loss": 0.5054,
+      "step": 218850
+    },
+    {
+      "epoch": 2.189,
+      "grad_norm": 90.06917572021484,
+      "learning_rate": 1.24756e-07,
+      "loss": 0.3561,
+      "step": 218900
+    },
+    {
+      "epoch": 2.1895,
+      "grad_norm": 0.39669129252433777,
+      "learning_rate": 1.24556e-07,
+      "loss": 0.4879,
+      "step": 218950
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 15.000761032104492,
+      "learning_rate": 1.2435599999999998e-07,
+      "loss": 0.5528,
+      "step": 219000
+    },
+    {
+      "epoch": 2.1905,
+      "grad_norm": 77.33452606201172,
+      "learning_rate": 1.24156e-07,
+      "loss": 0.593,
+      "step": 219050
+    },
+    {
+      "epoch": 2.191,
+      "grad_norm": 67.30721282958984,
+      "learning_rate": 1.23956e-07,
+      "loss": 0.4106,
+      "step": 219100
+    },
+    {
+      "epoch": 2.1915,
+      "grad_norm": 4.53434944152832,
+      "learning_rate": 1.2375599999999999e-07,
+      "loss": 0.5223,
+      "step": 219150
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 48.35958480834961,
+      "learning_rate": 1.23556e-07,
+      "loss": 0.362,
+      "step": 219200
+    },
+    {
+      "epoch": 2.1925,
+      "grad_norm": 82.18716430664062,
+      "learning_rate": 1.2335599999999998e-07,
+      "loss": 0.4678,
+      "step": 219250
+    },
+    {
+      "epoch": 2.193,
+      "grad_norm": 11.158646583557129,
+      "learning_rate": 1.23156e-07,
+      "loss": 0.3892,
+      "step": 219300
+    },
+    {
+      "epoch": 2.1935000000000002,
+      "grad_norm": 36.34510040283203,
+      "learning_rate": 1.2295599999999998e-07,
+      "loss": 0.3854,
+      "step": 219350
+    },
+    {
+      "epoch": 2.194,
+      "grad_norm": 24.300992965698242,
+      "learning_rate": 1.22756e-07,
+      "loss": 0.3248,
+      "step": 219400
+    },
+    {
+      "epoch": 2.1945,
+      "grad_norm": 56.57987976074219,
+      "learning_rate": 1.22556e-07,
+      "loss": 0.4331,
+      "step": 219450
+    },
+    {
+      "epoch": 2.195,
+      "grad_norm": 44.11698532104492,
+      "learning_rate": 1.2236e-07,
+      "loss": 0.3961,
+      "step": 219500
+    },
+    {
+      "epoch": 2.1955,
+      "grad_norm": 0.06106671318411827,
+      "learning_rate": 1.2216e-07,
+      "loss": 0.357,
+      "step": 219550
+    },
+    {
+      "epoch": 2.196,
+      "grad_norm": 22.567058563232422,
+      "learning_rate": 1.2196e-07,
+      "loss": 0.4296,
+      "step": 219600
+    },
+    {
+      "epoch": 2.1965,
+      "grad_norm": 95.61750030517578,
+      "learning_rate": 1.2176e-07,
+      "loss": 0.5135,
+      "step": 219650
+    },
+    {
+      "epoch": 2.197,
+      "grad_norm": 0.5937201976776123,
+      "learning_rate": 1.2155999999999998e-07,
+      "loss": 0.4198,
+      "step": 219700
+    },
+    {
+      "epoch": 2.1975,
+      "grad_norm": 9.500515937805176,
+      "learning_rate": 1.2136e-07,
+      "loss": 0.3876,
+      "step": 219750
+    },
+    {
+      "epoch": 2.198,
+      "grad_norm": 18.312498092651367,
+      "learning_rate": 1.2116e-07,
+      "loss": 0.5221,
+      "step": 219800
+    },
+    {
+      "epoch": 2.1985,
+      "grad_norm": 108.91061401367188,
+      "learning_rate": 1.2096e-07,
+      "loss": 0.4779,
+      "step": 219850
+    },
+    {
+      "epoch": 2.199,
+      "grad_norm": 108.51426696777344,
+      "learning_rate": 1.2076e-07,
+      "loss": 0.3632,
+      "step": 219900
+    },
+    {
+      "epoch": 2.1995,
+      "grad_norm": 54.85345458984375,
+      "learning_rate": 1.2056e-07,
+      "loss": 0.2912,
+      "step": 219950
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 98.45470428466797,
+      "learning_rate": 1.2036e-07,
+      "loss": 0.3563,
+      "step": 220000
+    },
+    {
+      "epoch": 2.2005,
+      "grad_norm": 26.045886993408203,
+      "learning_rate": 1.2015999999999999e-07,
+      "loss": 0.5571,
+      "step": 220050
+    },
+    {
+      "epoch": 2.201,
+      "grad_norm": 70.3372802734375,
+      "learning_rate": 1.1996e-07,
+      "loss": 0.4871,
+      "step": 220100
+    },
+    {
+      "epoch": 2.2015000000000002,
+      "grad_norm": 3.0247647762298584,
+      "learning_rate": 1.1976e-07,
+      "loss": 0.3335,
+      "step": 220150
+    },
+    {
+      "epoch": 2.202,
+      "grad_norm": 24.912132263183594,
+      "learning_rate": 1.1956e-07,
+      "loss": 0.4888,
+      "step": 220200
+    },
+    {
+      "epoch": 2.2025,
+      "grad_norm": 79.61896514892578,
+      "learning_rate": 1.1935999999999998e-07,
+      "loss": 0.5456,
+      "step": 220250
+    },
+    {
+      "epoch": 2.203,
+      "grad_norm": 52.392642974853516,
+      "learning_rate": 1.1916e-07,
+      "loss": 0.4036,
+      "step": 220300
+    },
+    {
+      "epoch": 2.2035,
+      "grad_norm": 2.282226085662842,
+      "learning_rate": 1.1895999999999999e-07,
+      "loss": 0.3552,
+      "step": 220350
+    },
+    {
+      "epoch": 2.204,
+      "grad_norm": 26.8724365234375,
+      "learning_rate": 1.1876e-07,
+      "loss": 0.3787,
+      "step": 220400
+    },
+    {
+      "epoch": 2.2045,
+      "grad_norm": 84.30389404296875,
+      "learning_rate": 1.1856e-07,
+      "loss": 0.3487,
+      "step": 220450
+    },
+    {
+      "epoch": 2.205,
+      "grad_norm": 74.3729476928711,
+      "learning_rate": 1.1836e-07,
+      "loss": 0.432,
+      "step": 220500
+    },
+    {
+      "epoch": 2.2055,
+      "grad_norm": 25.35677146911621,
+      "learning_rate": 1.1816e-07,
+      "loss": 0.4825,
+      "step": 220550
+    },
+    {
+      "epoch": 2.206,
+      "grad_norm": 5.5437116622924805,
+      "learning_rate": 1.1795999999999999e-07,
+      "loss": 0.4293,
+      "step": 220600
+    },
+    {
+      "epoch": 2.2065,
+      "grad_norm": 51.029869079589844,
+      "learning_rate": 1.1776e-07,
+      "loss": 0.3243,
+      "step": 220650
+    },
+    {
+      "epoch": 2.207,
+      "grad_norm": 12.300873756408691,
+      "learning_rate": 1.1755999999999999e-07,
+      "loss": 0.3076,
+      "step": 220700
+    },
+    {
+      "epoch": 2.2075,
+      "grad_norm": 0.5374252796173096,
+      "learning_rate": 1.1736e-07,
+      "loss": 0.4596,
+      "step": 220750
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 19.82467269897461,
+      "learning_rate": 1.1716e-07,
+      "loss": 0.4194,
+      "step": 220800
+    },
+    {
+      "epoch": 2.2085,
+      "grad_norm": 88.30901336669922,
+      "learning_rate": 1.1695999999999998e-07,
+      "loss": 0.4866,
+      "step": 220850
+    },
+    {
+      "epoch": 2.209,
+      "grad_norm": 53.6544303894043,
+      "learning_rate": 1.1676e-07,
+      "loss": 0.5025,
+      "step": 220900
+    },
+    {
+      "epoch": 2.2095,
+      "grad_norm": 16.769838333129883,
+      "learning_rate": 1.1655999999999999e-07,
+      "loss": 0.3527,
+      "step": 220950
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 77.25714874267578,
+      "learning_rate": 1.1636e-07,
+      "loss": 0.4571,
+      "step": 221000
+    },
+    {
+      "epoch": 2.2105,
+      "grad_norm": 74.47918701171875,
+      "learning_rate": 1.1615999999999999e-07,
+      "loss": 0.4852,
+      "step": 221050
+    },
+    {
+      "epoch": 2.211,
+      "grad_norm": 1.7386324405670166,
+      "learning_rate": 1.1595999999999999e-07,
+      "loss": 0.4368,
+      "step": 221100
+    },
+    {
+      "epoch": 2.2115,
+      "grad_norm": 43.10595703125,
+      "learning_rate": 1.1576e-07,
+      "loss": 0.472,
+      "step": 221150
+    },
+    {
+      "epoch": 2.212,
+      "grad_norm": 195.77264404296875,
+      "learning_rate": 1.1556e-07,
+      "loss": 0.4499,
+      "step": 221200
+    },
+    {
+      "epoch": 2.2125,
+      "grad_norm": 41.63266372680664,
+      "learning_rate": 1.1536e-07,
+      "loss": 0.3944,
+      "step": 221250
+    },
+    {
+      "epoch": 2.213,
+      "grad_norm": 93.48374938964844,
+      "learning_rate": 1.1516e-07,
+      "loss": 0.4368,
+      "step": 221300
+    },
+    {
+      "epoch": 2.2135,
+      "grad_norm": 11.143013000488281,
+      "learning_rate": 1.1496e-07,
+      "loss": 0.4911,
+      "step": 221350
+    },
+    {
+      "epoch": 2.214,
+      "grad_norm": 0.3506889045238495,
+      "learning_rate": 1.1475999999999999e-07,
+      "loss": 0.4476,
+      "step": 221400
+    },
+    {
+      "epoch": 2.2145,
+      "grad_norm": 5.332769393920898,
+      "learning_rate": 1.1455999999999999e-07,
+      "loss": 0.3238,
+      "step": 221450
+    },
+    {
+      "epoch": 2.215,
+      "grad_norm": 69.08348083496094,
+      "learning_rate": 1.1436e-07,
+      "loss": 0.4466,
+      "step": 221500
+    },
+    {
+      "epoch": 2.2155,
+      "grad_norm": 62.9998664855957,
+      "learning_rate": 1.1416e-07,
+      "loss": 0.4017,
+      "step": 221550
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 66.38433074951172,
+      "learning_rate": 1.1396e-07,
+      "loss": 0.4719,
+      "step": 221600
+    },
+    {
+      "epoch": 2.2165,
+      "grad_norm": 123.8575668334961,
+      "learning_rate": 1.1376e-07,
+      "loss": 0.3695,
+      "step": 221650
+    },
+    {
+      "epoch": 2.217,
+      "grad_norm": 9.564665794372559,
+      "learning_rate": 1.1355999999999999e-07,
+      "loss": 0.3205,
+      "step": 221700
+    },
+    {
+      "epoch": 2.2175,
+      "grad_norm": 28.491945266723633,
+      "learning_rate": 1.1335999999999999e-07,
+      "loss": 0.4048,
+      "step": 221750
+    },
+    {
+      "epoch": 2.218,
+      "grad_norm": 20.753578186035156,
+      "learning_rate": 1.1315999999999999e-07,
+      "loss": 0.4333,
+      "step": 221800
+    },
+    {
+      "epoch": 2.2185,
+      "grad_norm": 77.56809997558594,
+      "learning_rate": 1.1296e-07,
+      "loss": 0.5474,
+      "step": 221850
+    },
+    {
+      "epoch": 2.219,
+      "grad_norm": 6.686479091644287,
+      "learning_rate": 1.1276e-07,
+      "loss": 0.4709,
+      "step": 221900
+    },
+    {
+      "epoch": 2.2195,
+      "grad_norm": 1.914489507675171,
+      "learning_rate": 1.1255999999999998e-07,
+      "loss": 0.3998,
+      "step": 221950
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.9213142991065979,
+      "learning_rate": 1.1236e-07,
+      "loss": 0.378,
+      "step": 222000
+    },
+    {
+      "epoch": 2.2205,
+      "grad_norm": 115.55097198486328,
+      "learning_rate": 1.1215999999999999e-07,
+      "loss": 0.4008,
+      "step": 222050
+    },
+    {
+      "epoch": 2.221,
+      "grad_norm": 72.356201171875,
+      "learning_rate": 1.1196e-07,
+      "loss": 0.3891,
+      "step": 222100
+    },
+    {
+      "epoch": 2.2215,
+      "grad_norm": 6.992410659790039,
+      "learning_rate": 1.1175999999999999e-07,
+      "loss": 0.3579,
+      "step": 222150
+    },
+    {
+      "epoch": 2.222,
+      "grad_norm": 72.04354858398438,
+      "learning_rate": 1.1156e-07,
+      "loss": 0.402,
+      "step": 222200
+    },
+    {
+      "epoch": 2.2225,
+      "grad_norm": 112.33955383300781,
+      "learning_rate": 1.1136e-07,
+      "loss": 0.4516,
+      "step": 222250
+    },
+    {
+      "epoch": 2.223,
+      "grad_norm": 45.285400390625,
+      "learning_rate": 1.1115999999999998e-07,
+      "loss": 0.3904,
+      "step": 222300
+    },
+    {
+      "epoch": 2.2235,
+      "grad_norm": 0.01933535747230053,
+      "learning_rate": 1.1096e-07,
+      "loss": 0.4005,
+      "step": 222350
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 2.9697704315185547,
+      "learning_rate": 1.1076e-07,
+      "loss": 0.4953,
+      "step": 222400
+    },
+    {
+      "epoch": 2.2245,
+      "grad_norm": 143.373779296875,
+      "learning_rate": 1.1056e-07,
+      "loss": 0.3947,
+      "step": 222450
+    },
+    {
+      "epoch": 2.225,
+      "grad_norm": 2.298295259475708,
+      "learning_rate": 1.1035999999999999e-07,
+      "loss": 0.3572,
+      "step": 222500
+    },
+    {
+      "epoch": 2.2255,
+      "grad_norm": 4.240357875823975,
+      "learning_rate": 1.1015999999999999e-07,
+      "loss": 0.4634,
+      "step": 222550
+    },
+    {
+      "epoch": 2.226,
+      "grad_norm": 19.71664810180664,
+      "learning_rate": 1.0996e-07,
+      "loss": 0.4539,
+      "step": 222600
+    },
+    {
+      "epoch": 2.2265,
+      "grad_norm": 20.2246150970459,
+      "learning_rate": 1.0975999999999998e-07,
+      "loss": 0.4125,
+      "step": 222650
+    },
+    {
+      "epoch": 2.227,
+      "grad_norm": 9.409420013427734,
+      "learning_rate": 1.0956e-07,
+      "loss": 0.4532,
+      "step": 222700
+    },
+    {
+      "epoch": 2.2275,
+      "grad_norm": 2.0112721920013428,
+      "learning_rate": 1.0936e-07,
+      "loss": 0.4254,
+      "step": 222750
+    },
+    {
+      "epoch": 2.228,
+      "grad_norm": 107.7596664428711,
+      "learning_rate": 1.0915999999999999e-07,
+      "loss": 0.4481,
+      "step": 222800
+    },
+    {
+      "epoch": 2.2285,
+      "grad_norm": 64.18367767333984,
+      "learning_rate": 1.0895999999999999e-07,
+      "loss": 0.5081,
+      "step": 222850
+    },
+    {
+      "epoch": 2.229,
+      "grad_norm": 31.307912826538086,
+      "learning_rate": 1.0875999999999999e-07,
+      "loss": 0.411,
+      "step": 222900
+    },
+    {
+      "epoch": 2.2295,
+      "grad_norm": 62.03761291503906,
+      "learning_rate": 1.0856e-07,
+      "loss": 0.5153,
+      "step": 222950
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 68.89613342285156,
+      "learning_rate": 1.0836e-07,
+      "loss": 0.3965,
+      "step": 223000
+    },
+    {
+      "epoch": 2.2305,
+      "grad_norm": 10.568138122558594,
+      "learning_rate": 1.0816e-07,
+      "loss": 0.3033,
+      "step": 223050
+    },
+    {
+      "epoch": 2.231,
+      "grad_norm": 17.958087921142578,
+      "learning_rate": 1.07964e-07,
+      "loss": 0.428,
+      "step": 223100
+    },
+    {
+      "epoch": 2.2315,
+      "grad_norm": 1.1294639110565186,
+      "learning_rate": 1.07764e-07,
+      "loss": 0.4191,
+      "step": 223150
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 31.283260345458984,
+      "learning_rate": 1.07564e-07,
+      "loss": 0.3648,
+      "step": 223200
+    },
+    {
+      "epoch": 2.2325,
+      "grad_norm": 74.15252685546875,
+      "learning_rate": 1.07364e-07,
+      "loss": 0.3545,
+      "step": 223250
+    },
+    {
+      "epoch": 2.233,
+      "grad_norm": 56.10187911987305,
+      "learning_rate": 1.0716399999999999e-07,
+      "loss": 0.5046,
+      "step": 223300
+    },
+    {
+      "epoch": 2.2335,
+      "grad_norm": 19.8614444732666,
+      "learning_rate": 1.06964e-07,
+      "loss": 0.4121,
+      "step": 223350
+    },
+    {
+      "epoch": 2.234,
+      "grad_norm": 52.8697395324707,
+      "learning_rate": 1.0676399999999999e-07,
+      "loss": 0.4443,
+      "step": 223400
+    },
+    {
+      "epoch": 2.2345,
+      "grad_norm": 15.52030086517334,
+      "learning_rate": 1.06564e-07,
+      "loss": 0.4393,
+      "step": 223450
+    },
+    {
+      "epoch": 2.235,
+      "grad_norm": 13.279940605163574,
+      "learning_rate": 1.06364e-07,
+      "loss": 0.5055,
+      "step": 223500
+    },
+    {
+      "epoch": 2.2355,
+      "grad_norm": 35.77288818359375,
+      "learning_rate": 1.0616399999999998e-07,
+      "loss": 0.3935,
+      "step": 223550
+    },
+    {
+      "epoch": 2.2359999999999998,
+      "grad_norm": 41.148807525634766,
+      "learning_rate": 1.05964e-07,
+      "loss": 0.5071,
+      "step": 223600
+    },
+    {
+      "epoch": 2.2365,
+      "grad_norm": 119.54339599609375,
+      "learning_rate": 1.0576399999999999e-07,
+      "loss": 0.5073,
+      "step": 223650
+    },
+    {
+      "epoch": 2.237,
+      "grad_norm": 4.346285343170166,
+      "learning_rate": 1.05564e-07,
+      "loss": 0.3924,
+      "step": 223700
+    },
+    {
+      "epoch": 2.2375,
+      "grad_norm": 22.041217803955078,
+      "learning_rate": 1.0536399999999999e-07,
+      "loss": 0.4077,
+      "step": 223750
+    },
+    {
+      "epoch": 2.238,
+      "grad_norm": 23.932233810424805,
+      "learning_rate": 1.0516399999999999e-07,
+      "loss": 0.4078,
+      "step": 223800
+    },
+    {
+      "epoch": 2.2385,
+      "grad_norm": 92.13134765625,
+      "learning_rate": 1.04964e-07,
+      "loss": 0.3978,
+      "step": 223850
+    },
+    {
+      "epoch": 2.239,
+      "grad_norm": 5.41093635559082,
+      "learning_rate": 1.04764e-07,
+      "loss": 0.3965,
+      "step": 223900
+    },
+    {
+      "epoch": 2.2395,
+      "grad_norm": 32.597747802734375,
+      "learning_rate": 1.04564e-07,
+      "loss": 0.4309,
+      "step": 223950
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 76.27898406982422,
+      "learning_rate": 1.04364e-07,
+      "loss": 0.4534,
+      "step": 224000
+    },
+    {
+      "epoch": 2.2405,
+      "grad_norm": 77.87999725341797,
+      "learning_rate": 1.04164e-07,
+      "loss": 0.364,
+      "step": 224050
+    },
+    {
+      "epoch": 2.241,
+      "grad_norm": 83.9505615234375,
+      "learning_rate": 1.0396399999999999e-07,
+      "loss": 0.4332,
+      "step": 224100
+    },
+    {
+      "epoch": 2.2415,
+      "grad_norm": 92.03340148925781,
+      "learning_rate": 1.0376399999999999e-07,
+      "loss": 0.4405,
+      "step": 224150
+    },
+    {
+      "epoch": 2.242,
+      "grad_norm": 2.194648027420044,
+      "learning_rate": 1.03564e-07,
+      "loss": 0.5773,
+      "step": 224200
+    },
+    {
+      "epoch": 2.2425,
+      "grad_norm": 11.85653018951416,
+      "learning_rate": 1.03364e-07,
+      "loss": 0.2945,
+      "step": 224250
+    },
+    {
+      "epoch": 2.243,
+      "grad_norm": 77.39554595947266,
+      "learning_rate": 1.03164e-07,
+      "loss": 0.382,
+      "step": 224300
+    },
+    {
+      "epoch": 2.2435,
+      "grad_norm": 85.84123229980469,
+      "learning_rate": 1.02964e-07,
+      "loss": 0.3054,
+      "step": 224350
+    },
+    {
+      "epoch": 2.2439999999999998,
+      "grad_norm": 95.42002868652344,
+      "learning_rate": 1.0276399999999999e-07,
+      "loss": 0.4038,
+      "step": 224400
+    },
+    {
+      "epoch": 2.2445,
+      "grad_norm": 1.0675559043884277,
+      "learning_rate": 1.0256399999999999e-07,
+      "loss": 0.3923,
+      "step": 224450
+    },
+    {
+      "epoch": 2.245,
+      "grad_norm": 68.109130859375,
+      "learning_rate": 1.0236399999999999e-07,
+      "loss": 0.5414,
+      "step": 224500
+    },
+    {
+      "epoch": 2.2455,
+      "grad_norm": 14.229341506958008,
+      "learning_rate": 1.02164e-07,
+      "loss": 0.4831,
+      "step": 224550
+    },
+    {
+      "epoch": 2.246,
+      "grad_norm": 53.71510314941406,
+      "learning_rate": 1.01964e-07,
+      "loss": 0.4338,
+      "step": 224600
+    },
+    {
+      "epoch": 2.2465,
+      "grad_norm": 44.91522216796875,
+      "learning_rate": 1.0176399999999998e-07,
+      "loss": 0.4844,
+      "step": 224650
+    },
+    {
+      "epoch": 2.247,
+      "grad_norm": 138.286865234375,
+      "learning_rate": 1.01564e-07,
+      "loss": 0.361,
+      "step": 224700
+    },
+    {
+      "epoch": 2.2475,
+      "grad_norm": 43.51007080078125,
+      "learning_rate": 1.01364e-07,
+      "loss": 0.4108,
+      "step": 224750
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 7.015733242034912,
+      "learning_rate": 1.01164e-07,
+      "loss": 0.3845,
+      "step": 224800
+    },
+    {
+      "epoch": 2.2485,
+      "grad_norm": 125.538330078125,
+      "learning_rate": 1.0096399999999999e-07,
+      "loss": 0.463,
+      "step": 224850
+    },
+    {
+      "epoch": 2.249,
+      "grad_norm": 7.6888346672058105,
+      "learning_rate": 1.00764e-07,
+      "loss": 0.4631,
+      "step": 224900
+    },
+    {
+      "epoch": 2.2495,
+      "grad_norm": 6.268250465393066,
+      "learning_rate": 1.00564e-07,
+      "loss": 0.2814,
+      "step": 224950
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.0146398544311523,
+      "learning_rate": 1.0036399999999998e-07,
+      "loss": 0.4237,
+      "step": 225000
+    },
+    {
+      "epoch": 2.2505,
+      "grad_norm": 33.02668380737305,
+      "learning_rate": 1.00164e-07,
+      "loss": 0.3576,
+      "step": 225050
+    },
+    {
+      "epoch": 2.251,
+      "grad_norm": 97.76544189453125,
+      "learning_rate": 9.9964e-08,
+      "loss": 0.3949,
+      "step": 225100
+    },
+    {
+      "epoch": 2.2515,
+      "grad_norm": 78.87450408935547,
+      "learning_rate": 9.9764e-08,
+      "loss": 0.2917,
+      "step": 225150
+    },
+    {
+      "epoch": 2.252,
+      "grad_norm": 108.28045654296875,
+      "learning_rate": 9.956399999999999e-08,
+      "loss": 0.5591,
+      "step": 225200
+    },
+    {
+      "epoch": 2.2525,
+      "grad_norm": 65.84611511230469,
+      "learning_rate": 9.936399999999999e-08,
+      "loss": 0.3784,
+      "step": 225250
+    },
+    {
+      "epoch": 2.253,
+      "grad_norm": 20.012842178344727,
+      "learning_rate": 9.9164e-08,
+      "loss": 0.4402,
+      "step": 225300
+    },
+    {
+      "epoch": 2.2535,
+      "grad_norm": 1.915622353553772,
+      "learning_rate": 9.896399999999999e-08,
+      "loss": 0.4123,
+      "step": 225350
+    },
+    {
+      "epoch": 2.254,
+      "grad_norm": 19.519975662231445,
+      "learning_rate": 9.8764e-08,
+      "loss": 0.3652,
+      "step": 225400
+    },
+    {
+      "epoch": 2.2545,
+      "grad_norm": 87.41107940673828,
+      "learning_rate": 9.8564e-08,
+      "loss": 0.6044,
+      "step": 225450
+    },
+    {
+      "epoch": 2.255,
+      "grad_norm": 29.000362396240234,
+      "learning_rate": 9.836399999999999e-08,
+      "loss": 0.4545,
+      "step": 225500
+    },
+    {
+      "epoch": 2.2555,
+      "grad_norm": 70.12849426269531,
+      "learning_rate": 9.816399999999999e-08,
+      "loss": 0.4011,
+      "step": 225550
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.5577579140663147,
+      "learning_rate": 9.796399999999999e-08,
+      "loss": 0.4326,
+      "step": 225600
+    },
+    {
+      "epoch": 2.2565,
+      "grad_norm": 2.2879648208618164,
+      "learning_rate": 9.7764e-08,
+      "loss": 0.3948,
+      "step": 225650
+    },
+    {
+      "epoch": 2.257,
+      "grad_norm": 0.7911091446876526,
+      "learning_rate": 9.7564e-08,
+      "loss": 0.5271,
+      "step": 225700
+    },
+    {
+      "epoch": 2.2575,
+      "grad_norm": 52.76624298095703,
+      "learning_rate": 9.7364e-08,
+      "loss": 0.5534,
+      "step": 225750
+    },
+    {
+      "epoch": 2.258,
+      "grad_norm": 53.3375244140625,
+      "learning_rate": 9.7164e-08,
+      "loss": 0.4334,
+      "step": 225800
+    },
+    {
+      "epoch": 2.2585,
+      "grad_norm": 27.924394607543945,
+      "learning_rate": 9.6964e-08,
+      "loss": 0.2814,
+      "step": 225850
+    },
+    {
+      "epoch": 2.259,
+      "grad_norm": 48.85221481323242,
+      "learning_rate": 9.676399999999999e-08,
+      "loss": 0.3435,
+      "step": 225900
+    },
+    {
+      "epoch": 2.2595,
+      "grad_norm": 7.6830549240112305,
+      "learning_rate": 9.656399999999999e-08,
+      "loss": 0.5507,
+      "step": 225950
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.447178840637207,
+      "learning_rate": 9.6364e-08,
+      "loss": 0.329,
+      "step": 226000
+    },
+    {
+      "epoch": 2.2605,
+      "grad_norm": 39.04615783691406,
+      "learning_rate": 9.6164e-08,
+      "loss": 0.4663,
+      "step": 226050
+    },
+    {
+      "epoch": 2.261,
+      "grad_norm": 4.852475643157959,
+      "learning_rate": 9.596399999999999e-08,
+      "loss": 0.5203,
+      "step": 226100
+    },
+    {
+      "epoch": 2.2615,
+      "grad_norm": 89.7945556640625,
+      "learning_rate": 9.5764e-08,
+      "loss": 0.4516,
+      "step": 226150
+    },
+    {
+      "epoch": 2.262,
+      "grad_norm": 35.014644622802734,
+      "learning_rate": 9.5564e-08,
+      "loss": 0.3496,
+      "step": 226200
+    },
+    {
+      "epoch": 2.2625,
+      "grad_norm": 57.93213653564453,
+      "learning_rate": 9.5364e-08,
+      "loss": 0.4799,
+      "step": 226250
+    },
+    {
+      "epoch": 2.263,
+      "grad_norm": 39.42817306518555,
+      "learning_rate": 9.516399999999999e-08,
+      "loss": 0.4357,
+      "step": 226300
+    },
+    {
+      "epoch": 2.2635,
+      "grad_norm": 41.9278450012207,
+      "learning_rate": 9.4964e-08,
+      "loss": 0.437,
+      "step": 226350
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 41.00620651245117,
+      "learning_rate": 9.4764e-08,
+      "loss": 0.417,
+      "step": 226400
+    },
+    {
+      "epoch": 2.2645,
+      "grad_norm": 90.83309936523438,
+      "learning_rate": 9.456399999999999e-08,
+      "loss": 0.4885,
+      "step": 226450
+    },
+    {
+      "epoch": 2.265,
+      "grad_norm": 36.397216796875,
+      "learning_rate": 9.4364e-08,
+      "loss": 0.4193,
+      "step": 226500
+    },
+    {
+      "epoch": 2.2655,
+      "grad_norm": 43.22880554199219,
+      "learning_rate": 9.4164e-08,
+      "loss": 0.4173,
+      "step": 226550
+    },
+    {
+      "epoch": 2.266,
+      "grad_norm": 144.54409790039062,
+      "learning_rate": 9.396400000000001e-08,
+      "loss": 0.3163,
+      "step": 226600
+    },
+    {
+      "epoch": 2.2665,
+      "grad_norm": 0.3794969916343689,
+      "learning_rate": 9.376399999999999e-08,
+      "loss": 0.3984,
+      "step": 226650
+    },
+    {
+      "epoch": 2.267,
+      "grad_norm": 48.13538360595703,
+      "learning_rate": 9.356399999999999e-08,
+      "loss": 0.3953,
+      "step": 226700
+    },
+    {
+      "epoch": 2.2675,
+      "grad_norm": 89.7881088256836,
+      "learning_rate": 9.3364e-08,
+      "loss": 0.5402,
+      "step": 226750
+    },
+    {
+      "epoch": 2.268,
+      "grad_norm": 87.7471923828125,
+      "learning_rate": 9.316399999999999e-08,
+      "loss": 0.438,
+      "step": 226800
+    },
+    {
+      "epoch": 2.2685,
+      "grad_norm": 22.916074752807617,
+      "learning_rate": 9.2964e-08,
+      "loss": 0.3621,
+      "step": 226850
+    },
+    {
+      "epoch": 2.269,
+      "grad_norm": 23.842510223388672,
+      "learning_rate": 9.2764e-08,
+      "loss": 0.351,
+      "step": 226900
+    },
+    {
+      "epoch": 2.2695,
+      "grad_norm": 86.34049224853516,
+      "learning_rate": 9.2564e-08,
+      "loss": 0.4619,
+      "step": 226950
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 34.528133392333984,
+      "learning_rate": 9.236399999999999e-08,
+      "loss": 0.367,
+      "step": 227000
+    },
+    {
+      "epoch": 2.2705,
+      "grad_norm": 42.39723205566406,
+      "learning_rate": 9.216399999999999e-08,
+      "loss": 0.4345,
+      "step": 227050
+    },
+    {
+      "epoch": 2.271,
+      "grad_norm": 4.247285842895508,
+      "learning_rate": 9.1964e-08,
+      "loss": 0.6515,
+      "step": 227100
+    },
+    {
+      "epoch": 2.2715,
+      "grad_norm": 34.562137603759766,
+      "learning_rate": 9.1764e-08,
+      "loss": 0.4751,
+      "step": 227150
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 0.823958158493042,
+      "learning_rate": 9.1564e-08,
+      "loss": 0.4003,
+      "step": 227200
+    },
+    {
+      "epoch": 2.2725,
+      "grad_norm": 86.37012481689453,
+      "learning_rate": 9.1364e-08,
+      "loss": 0.4309,
+      "step": 227250
+    },
+    {
+      "epoch": 2.273,
+      "grad_norm": 108.56483459472656,
+      "learning_rate": 9.1164e-08,
+      "loss": 0.5015,
+      "step": 227300
+    },
+    {
+      "epoch": 2.2735,
+      "grad_norm": 73.98970031738281,
+      "learning_rate": 9.096399999999999e-08,
+      "loss": 0.3684,
+      "step": 227350
+    },
+    {
+      "epoch": 2.274,
+      "grad_norm": 18.87233543395996,
+      "learning_rate": 9.076399999999999e-08,
+      "loss": 0.4713,
+      "step": 227400
+    },
+    {
+      "epoch": 2.2745,
+      "grad_norm": 88.28274536132812,
+      "learning_rate": 9.0564e-08,
+      "loss": 0.5386,
+      "step": 227450
+    },
+    {
+      "epoch": 2.275,
+      "grad_norm": 63.17116165161133,
+      "learning_rate": 9.0364e-08,
+      "loss": 0.4483,
+      "step": 227500
+    },
+    {
+      "epoch": 2.2755,
+      "grad_norm": 28.934703826904297,
+      "learning_rate": 9.016399999999999e-08,
+      "loss": 0.4424,
+      "step": 227550
+    },
+    {
+      "epoch": 2.276,
+      "grad_norm": 61.9217414855957,
+      "learning_rate": 8.9964e-08,
+      "loss": 0.4156,
+      "step": 227600
+    },
+    {
+      "epoch": 2.2765,
+      "grad_norm": 75.80464172363281,
+      "learning_rate": 8.9764e-08,
+      "loss": 0.4462,
+      "step": 227650
+    },
+    {
+      "epoch": 2.277,
+      "grad_norm": 9.887880325317383,
+      "learning_rate": 8.956400000000001e-08,
+      "loss": 0.4142,
+      "step": 227700
+    },
+    {
+      "epoch": 2.2775,
+      "grad_norm": 46.1187858581543,
+      "learning_rate": 8.936399999999999e-08,
+      "loss": 0.3884,
+      "step": 227750
+    },
+    {
+      "epoch": 2.278,
+      "grad_norm": 42.99738311767578,
+      "learning_rate": 8.916399999999999e-08,
+      "loss": 0.4486,
+      "step": 227800
+    },
+    {
+      "epoch": 2.2785,
+      "grad_norm": 5.804699420928955,
+      "learning_rate": 8.8968e-08,
+      "loss": 0.3833,
+      "step": 227850
+    },
+    {
+      "epoch": 2.279,
+      "grad_norm": 20.068557739257812,
+      "learning_rate": 8.876799999999999e-08,
+      "loss": 0.4945,
+      "step": 227900
+    },
+    {
+      "epoch": 2.2795,
+      "grad_norm": 0.5861179828643799,
+      "learning_rate": 8.856799999999999e-08,
+      "loss": 0.4478,
+      "step": 227950
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.990058958530426,
+      "learning_rate": 8.8368e-08,
+      "loss": 0.3093,
+      "step": 228000
+    },
+    {
+      "epoch": 2.2805,
+      "grad_norm": 129.85740661621094,
+      "learning_rate": 8.816799999999999e-08,
+      "loss": 0.429,
+      "step": 228050
+    },
+    {
+      "epoch": 2.281,
+      "grad_norm": 29.716796875,
+      "learning_rate": 8.7968e-08,
+      "loss": 0.3996,
+      "step": 228100
+    },
+    {
+      "epoch": 2.2815,
+      "grad_norm": 58.8203239440918,
+      "learning_rate": 8.7768e-08,
+      "loss": 0.3056,
+      "step": 228150
+    },
+    {
+      "epoch": 2.282,
+      "grad_norm": 41.287166595458984,
+      "learning_rate": 8.7568e-08,
+      "loss": 0.321,
+      "step": 228200
+    },
+    {
+      "epoch": 2.2824999999999998,
+      "grad_norm": 9.633661270141602,
+      "learning_rate": 8.736799999999999e-08,
+      "loss": 0.52,
+      "step": 228250
+    },
+    {
+      "epoch": 2.283,
+      "grad_norm": 35.10003662109375,
+      "learning_rate": 8.716799999999999e-08,
+      "loss": 0.4273,
+      "step": 228300
+    },
+    {
+      "epoch": 2.2835,
+      "grad_norm": 8.269124031066895,
+      "learning_rate": 8.6968e-08,
+      "loss": 0.3968,
+      "step": 228350
+    },
+    {
+      "epoch": 2.284,
+      "grad_norm": 47.246002197265625,
+      "learning_rate": 8.6768e-08,
+      "loss": 0.6276,
+      "step": 228400
+    },
+    {
+      "epoch": 2.2845,
+      "grad_norm": 106.63582611083984,
+      "learning_rate": 8.6568e-08,
+      "loss": 0.3303,
+      "step": 228450
+    },
+    {
+      "epoch": 2.285,
+      "grad_norm": 5.064298152923584,
+      "learning_rate": 8.6368e-08,
+      "loss": 0.46,
+      "step": 228500
+    },
+    {
+      "epoch": 2.2855,
+      "grad_norm": 0.2619529962539673,
+      "learning_rate": 8.6168e-08,
+      "loss": 0.3292,
+      "step": 228550
+    },
+    {
+      "epoch": 2.286,
+      "grad_norm": 7.505974292755127,
+      "learning_rate": 8.596799999999999e-08,
+      "loss": 0.4411,
+      "step": 228600
+    },
+    {
+      "epoch": 2.2865,
+      "grad_norm": 69.65245056152344,
+      "learning_rate": 8.576799999999999e-08,
+      "loss": 0.4029,
+      "step": 228650
+    },
+    {
+      "epoch": 2.287,
+      "grad_norm": 57.54338073730469,
+      "learning_rate": 8.5568e-08,
+      "loss": 0.4123,
+      "step": 228700
+    },
+    {
+      "epoch": 2.2875,
+      "grad_norm": 78.23273468017578,
+      "learning_rate": 8.5368e-08,
+      "loss": 0.3677,
+      "step": 228750
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 43.50226593017578,
+      "learning_rate": 8.516799999999999e-08,
+      "loss": 0.4841,
+      "step": 228800
+    },
+    {
+      "epoch": 2.2885,
+      "grad_norm": 160.9630126953125,
+      "learning_rate": 8.4968e-08,
+      "loss": 0.3561,
+      "step": 228850
+    },
+    {
+      "epoch": 2.289,
+      "grad_norm": 106.46874237060547,
+      "learning_rate": 8.4768e-08,
+      "loss": 0.3973,
+      "step": 228900
+    },
+    {
+      "epoch": 2.2895,
+      "grad_norm": 67.45094299316406,
+      "learning_rate": 8.4568e-08,
+      "loss": 0.498,
+      "step": 228950
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 100.94178771972656,
+      "learning_rate": 8.436799999999999e-08,
+      "loss": 0.542,
+      "step": 229000
+    },
+    {
+      "epoch": 2.2904999999999998,
+      "grad_norm": 3.8312125205993652,
+      "learning_rate": 8.4168e-08,
+      "loss": 0.7298,
+      "step": 229050
+    },
+    {
+      "epoch": 2.291,
+      "grad_norm": 100.0070571899414,
+      "learning_rate": 8.3968e-08,
+      "loss": 0.4864,
+      "step": 229100
+    },
+    {
+      "epoch": 2.2915,
+      "grad_norm": 88.91947937011719,
+      "learning_rate": 8.376799999999999e-08,
+      "loss": 0.4731,
+      "step": 229150
+    },
+    {
+      "epoch": 2.292,
+      "grad_norm": 11.505712509155273,
+      "learning_rate": 8.3568e-08,
+      "loss": 0.42,
+      "step": 229200
+    },
+    {
+      "epoch": 2.2925,
+      "grad_norm": 50.03477478027344,
+      "learning_rate": 8.3368e-08,
+      "loss": 0.433,
+      "step": 229250
+    },
+    {
+      "epoch": 2.293,
+      "grad_norm": 62.71595001220703,
+      "learning_rate": 8.316800000000001e-08,
+      "loss": 0.4322,
+      "step": 229300
+    },
+    {
+      "epoch": 2.2935,
+      "grad_norm": 81.32498931884766,
+      "learning_rate": 8.296799999999999e-08,
+      "loss": 0.4962,
+      "step": 229350
+    },
+    {
+      "epoch": 2.294,
+      "grad_norm": 80.8553466796875,
+      "learning_rate": 8.276799999999999e-08,
+      "loss": 0.3587,
+      "step": 229400
+    },
+    {
+      "epoch": 2.2945,
+      "grad_norm": 77.47965240478516,
+      "learning_rate": 8.2568e-08,
+      "loss": 0.3187,
+      "step": 229450
+    },
+    {
+      "epoch": 2.295,
+      "grad_norm": 77.37553405761719,
+      "learning_rate": 8.236799999999999e-08,
+      "loss": 0.5271,
+      "step": 229500
+    },
+    {
+      "epoch": 2.2955,
+      "grad_norm": 0.49823853373527527,
+      "learning_rate": 8.2168e-08,
+      "loss": 0.3428,
+      "step": 229550
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 2.033798933029175,
+      "learning_rate": 8.1968e-08,
+      "loss": 0.32,
+      "step": 229600
+    },
+    {
+      "epoch": 2.2965,
+      "grad_norm": 2.363724946975708,
+      "learning_rate": 8.1768e-08,
+      "loss": 0.2763,
+      "step": 229650
+    },
+    {
+      "epoch": 2.297,
+      "grad_norm": 42.67679977416992,
+      "learning_rate": 8.156799999999999e-08,
+      "loss": 0.3813,
+      "step": 229700
+    },
+    {
+      "epoch": 2.2975,
+      "grad_norm": 93.03001403808594,
+      "learning_rate": 8.1372e-08,
+      "loss": 0.4763,
+      "step": 229750
+    },
+    {
+      "epoch": 2.298,
+      "grad_norm": 23.316856384277344,
+      "learning_rate": 8.1172e-08,
+      "loss": 0.3295,
+      "step": 229800
+    },
+    {
+      "epoch": 2.2984999999999998,
+      "grad_norm": 57.193267822265625,
+      "learning_rate": 8.097199999999999e-08,
+      "loss": 0.5061,
+      "step": 229850
+    },
+    {
+      "epoch": 2.299,
+      "grad_norm": 22.44532585144043,
+      "learning_rate": 8.077199999999999e-08,
+      "loss": 0.3112,
+      "step": 229900
+    },
+    {
+      "epoch": 2.2995,
+      "grad_norm": 101.59669494628906,
+      "learning_rate": 8.0572e-08,
+      "loss": 0.3241,
+      "step": 229950
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 78.62955474853516,
+      "learning_rate": 8.0372e-08,
+      "loss": 0.4214,
+      "step": 230000
+    },
+    {
+      "epoch": 2.3005,
+      "grad_norm": 55.72152328491211,
+      "learning_rate": 8.017199999999998e-08,
+      "loss": 0.3366,
+      "step": 230050
+    },
+    {
+      "epoch": 2.301,
+      "grad_norm": 105.1781997680664,
+      "learning_rate": 7.9972e-08,
+      "loss": 0.5508,
+      "step": 230100
+    },
+    {
+      "epoch": 2.3015,
+      "grad_norm": 32.745849609375,
+      "learning_rate": 7.9772e-08,
+      "loss": 0.3962,
+      "step": 230150
+    },
+    {
+      "epoch": 2.302,
+      "grad_norm": 4.188356399536133,
+      "learning_rate": 7.9572e-08,
+      "loss": 0.3405,
+      "step": 230200
+    },
+    {
+      "epoch": 2.3025,
+      "grad_norm": 86.73455047607422,
+      "learning_rate": 7.937199999999999e-08,
+      "loss": 0.309,
+      "step": 230250
+    },
+    {
+      "epoch": 2.303,
+      "grad_norm": 96.70193481445312,
+      "learning_rate": 7.9172e-08,
+      "loss": 0.3716,
+      "step": 230300
+    },
+    {
+      "epoch": 2.3035,
+      "grad_norm": 2.2225139141082764,
+      "learning_rate": 7.8972e-08,
+      "loss": 0.3604,
+      "step": 230350
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 105.9201431274414,
+      "learning_rate": 7.877199999999999e-08,
+      "loss": 0.4714,
+      "step": 230400
+    },
+    {
+      "epoch": 2.3045,
+      "grad_norm": 83.0341567993164,
+      "learning_rate": 7.8572e-08,
+      "loss": 0.5562,
+      "step": 230450
+    },
+    {
+      "epoch": 2.305,
+      "grad_norm": 2.381653070449829,
+      "learning_rate": 7.8372e-08,
+      "loss": 0.4385,
+      "step": 230500
+    },
+    {
+      "epoch": 2.3055,
+      "grad_norm": 18.078433990478516,
+      "learning_rate": 7.817200000000001e-08,
+      "loss": 0.379,
+      "step": 230550
+    },
+    {
+      "epoch": 2.306,
+      "grad_norm": 27.159372329711914,
+      "learning_rate": 7.797199999999999e-08,
+      "loss": 0.3722,
+      "step": 230600
+    },
+    {
+      "epoch": 2.3064999999999998,
+      "grad_norm": 67.36116027832031,
+      "learning_rate": 7.777199999999999e-08,
+      "loss": 0.4527,
+      "step": 230650
+    },
+    {
+      "epoch": 2.307,
+      "grad_norm": 54.01509475708008,
+      "learning_rate": 7.7572e-08,
+      "loss": 0.3647,
+      "step": 230700
+    },
+    {
+      "epoch": 2.3075,
+      "grad_norm": 3.8897550106048584,
+      "learning_rate": 7.737199999999999e-08,
+      "loss": 0.5743,
+      "step": 230750
+    },
+    {
+      "epoch": 2.308,
+      "grad_norm": 0.7189392447471619,
+      "learning_rate": 7.7172e-08,
+      "loss": 0.4053,
+      "step": 230800
+    },
+    {
+      "epoch": 2.3085,
+      "grad_norm": 53.79637908935547,
+      "learning_rate": 7.6972e-08,
+      "loss": 0.6112,
+      "step": 230850
+    },
+    {
+      "epoch": 2.309,
+      "grad_norm": 29.953990936279297,
+      "learning_rate": 7.677200000000001e-08,
+      "loss": 0.4461,
+      "step": 230900
+    },
+    {
+      "epoch": 2.3095,
+      "grad_norm": 0.23817132413387299,
+      "learning_rate": 7.657199999999999e-08,
+      "loss": 0.4433,
+      "step": 230950
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 72.25556945800781,
+      "learning_rate": 7.637199999999999e-08,
+      "loss": 0.3299,
+      "step": 231000
+    },
+    {
+      "epoch": 2.3105,
+      "grad_norm": 56.55588150024414,
+      "learning_rate": 7.6172e-08,
+      "loss": 0.4597,
+      "step": 231050
+    },
+    {
+      "epoch": 2.311,
+      "grad_norm": 94.5429916381836,
+      "learning_rate": 7.5972e-08,
+      "loss": 0.3237,
+      "step": 231100
+    },
+    {
+      "epoch": 2.3115,
+      "grad_norm": 65.64086151123047,
+      "learning_rate": 7.5772e-08,
+      "loss": 0.4391,
+      "step": 231150
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 57.61141586303711,
+      "learning_rate": 7.5572e-08,
+      "loss": 0.3361,
+      "step": 231200
+    },
+    {
+      "epoch": 2.3125,
+      "grad_norm": 73.47406768798828,
+      "learning_rate": 7.5372e-08,
+      "loss": 0.4478,
+      "step": 231250
+    },
+    {
+      "epoch": 2.313,
+      "grad_norm": 8.852334976196289,
+      "learning_rate": 7.517199999999999e-08,
+      "loss": 0.56,
+      "step": 231300
+    },
+    {
+      "epoch": 2.3135,
+      "grad_norm": 86.10330200195312,
+      "learning_rate": 7.497199999999999e-08,
+      "loss": 0.3869,
+      "step": 231350
+    },
+    {
+      "epoch": 2.314,
+      "grad_norm": 26.84288787841797,
+      "learning_rate": 7.4772e-08,
+      "loss": 0.4664,
+      "step": 231400
+    },
+    {
+      "epoch": 2.3145,
+      "grad_norm": 28.752120971679688,
+      "learning_rate": 7.4572e-08,
+      "loss": 0.4008,
+      "step": 231450
+    },
+    {
+      "epoch": 2.315,
+      "grad_norm": 12.051288604736328,
+      "learning_rate": 7.437199999999999e-08,
+      "loss": 0.4183,
+      "step": 231500
+    },
+    {
+      "epoch": 2.3155,
+      "grad_norm": 60.44932556152344,
+      "learning_rate": 7.4172e-08,
+      "loss": 0.3753,
+      "step": 231550
+    },
+    {
+      "epoch": 2.316,
+      "grad_norm": 105.76837158203125,
+      "learning_rate": 7.3972e-08,
+      "loss": 0.4456,
+      "step": 231600
+    },
+    {
+      "epoch": 2.3165,
+      "grad_norm": 20.609785079956055,
+      "learning_rate": 7.377200000000001e-08,
+      "loss": 0.3352,
+      "step": 231650
+    },
+    {
+      "epoch": 2.317,
+      "grad_norm": 23.5540828704834,
+      "learning_rate": 7.357199999999999e-08,
+      "loss": 0.4725,
+      "step": 231700
+    },
+    {
+      "epoch": 2.3175,
+      "grad_norm": 9.055368423461914,
+      "learning_rate": 7.3372e-08,
+      "loss": 0.3954,
+      "step": 231750
+    },
+    {
+      "epoch": 2.318,
+      "grad_norm": 0.28335410356521606,
+      "learning_rate": 7.3172e-08,
+      "loss": 0.3075,
+      "step": 231800
+    },
+    {
+      "epoch": 2.3185000000000002,
+      "grad_norm": 142.01795959472656,
+      "learning_rate": 7.297199999999999e-08,
+      "loss": 0.454,
+      "step": 231850
+    },
+    {
+      "epoch": 2.319,
+      "grad_norm": 69.494873046875,
+      "learning_rate": 7.2772e-08,
+      "loss": 0.3512,
+      "step": 231900
+    },
+    {
+      "epoch": 2.3195,
+      "grad_norm": 144.35302734375,
+      "learning_rate": 7.2572e-08,
+      "loss": 0.3771,
+      "step": 231950
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 44.31450653076172,
+      "learning_rate": 7.237200000000001e-08,
+      "loss": 0.3828,
+      "step": 232000
+    },
+    {
+      "epoch": 2.3205,
+      "grad_norm": 43.578590393066406,
+      "learning_rate": 7.217199999999999e-08,
+      "loss": 0.4416,
+      "step": 232050
+    },
+    {
+      "epoch": 2.321,
+      "grad_norm": 0.5683667659759521,
+      "learning_rate": 7.197199999999999e-08,
+      "loss": 0.3824,
+      "step": 232100
+    },
+    {
+      "epoch": 2.3215,
+      "grad_norm": 13.191206932067871,
+      "learning_rate": 7.1772e-08,
+      "loss": 0.4271,
+      "step": 232150
+    },
+    {
+      "epoch": 2.322,
+      "grad_norm": 89.80245971679688,
+      "learning_rate": 7.157199999999999e-08,
+      "loss": 0.3694,
+      "step": 232200
+    },
+    {
+      "epoch": 2.3225,
+      "grad_norm": 100.60424041748047,
+      "learning_rate": 7.1372e-08,
+      "loss": 0.4274,
+      "step": 232250
+    },
+    {
+      "epoch": 2.323,
+      "grad_norm": 21.0079288482666,
+      "learning_rate": 7.1172e-08,
+      "loss": 0.4433,
+      "step": 232300
+    },
+    {
+      "epoch": 2.3235,
+      "grad_norm": 80.20172882080078,
+      "learning_rate": 7.0972e-08,
+      "loss": 0.2979,
+      "step": 232350
+    },
+    {
+      "epoch": 2.324,
+      "grad_norm": 22.903003692626953,
+      "learning_rate": 7.0772e-08,
+      "loss": 0.3579,
+      "step": 232400
+    },
+    {
+      "epoch": 2.3245,
+      "grad_norm": 69.52226257324219,
+      "learning_rate": 7.057199999999999e-08,
+      "loss": 0.524,
+      "step": 232450
+    },
+    {
+      "epoch": 2.325,
+      "grad_norm": 54.36795425415039,
+      "learning_rate": 7.0372e-08,
+      "loss": 0.5302,
+      "step": 232500
+    },
+    {
+      "epoch": 2.3255,
+      "grad_norm": 3.2600345611572266,
+      "learning_rate": 7.0172e-08,
+      "loss": 0.3502,
+      "step": 232550
+    },
+    {
+      "epoch": 2.326,
+      "grad_norm": 8.641088485717773,
+      "learning_rate": 6.9972e-08,
+      "loss": 0.4258,
+      "step": 232600
+    },
+    {
+      "epoch": 2.3265000000000002,
+      "grad_norm": 0.2351902574300766,
+      "learning_rate": 6.9772e-08,
+      "loss": 0.4205,
+      "step": 232650
+    },
+    {
+      "epoch": 2.327,
+      "grad_norm": 5.2298126220703125,
+      "learning_rate": 6.9572e-08,
+      "loss": 0.2792,
+      "step": 232700
+    },
+    {
+      "epoch": 2.3275,
+      "grad_norm": 120.54157257080078,
+      "learning_rate": 6.9372e-08,
+      "loss": 0.3501,
+      "step": 232750
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 129.25265502929688,
+      "learning_rate": 6.917199999999999e-08,
+      "loss": 0.5156,
+      "step": 232800
+    },
+    {
+      "epoch": 2.3285,
+      "grad_norm": 66.50567626953125,
+      "learning_rate": 6.8972e-08,
+      "loss": 0.5007,
+      "step": 232850
+    },
+    {
+      "epoch": 2.329,
+      "grad_norm": 80.42890930175781,
+      "learning_rate": 6.8772e-08,
+      "loss": 0.4191,
+      "step": 232900
+    },
+    {
+      "epoch": 2.3295,
+      "grad_norm": 84.22823333740234,
+      "learning_rate": 6.857199999999999e-08,
+      "loss": 0.4093,
+      "step": 232950
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 60.78300094604492,
+      "learning_rate": 6.8372e-08,
+      "loss": 0.3831,
+      "step": 233000
+    },
+    {
+      "epoch": 2.3305,
+      "grad_norm": 9.814701080322266,
+      "learning_rate": 6.8172e-08,
+      "loss": 0.3795,
+      "step": 233050
+    },
+    {
+      "epoch": 2.331,
+      "grad_norm": 89.87799072265625,
+      "learning_rate": 6.797200000000001e-08,
+      "loss": 0.4236,
+      "step": 233100
+    },
+    {
+      "epoch": 2.3315,
+      "grad_norm": 51.194068908691406,
+      "learning_rate": 6.777199999999999e-08,
+      "loss": 0.3,
+      "step": 233150
+    },
+    {
+      "epoch": 2.332,
+      "grad_norm": 40.99766159057617,
+      "learning_rate": 6.757199999999999e-08,
+      "loss": 0.5982,
+      "step": 233200
+    },
+    {
+      "epoch": 2.3325,
+      "grad_norm": 97.21105194091797,
+      "learning_rate": 6.7372e-08,
+      "loss": 0.4268,
+      "step": 233250
+    },
+    {
+      "epoch": 2.333,
+      "grad_norm": 77.22876739501953,
+      "learning_rate": 6.717199999999999e-08,
+      "loss": 0.3962,
+      "step": 233300
+    },
+    {
+      "epoch": 2.3335,
+      "grad_norm": 24.11876106262207,
+      "learning_rate": 6.6972e-08,
+      "loss": 0.3584,
+      "step": 233350
+    },
+    {
+      "epoch": 2.334,
+      "grad_norm": 49.48302459716797,
+      "learning_rate": 6.6772e-08,
+      "loss": 0.3827,
+      "step": 233400
+    },
+    {
+      "epoch": 2.3345000000000002,
+      "grad_norm": 0.11236397176980972,
+      "learning_rate": 6.657200000000001e-08,
+      "loss": 0.2988,
+      "step": 233450
+    },
+    {
+      "epoch": 2.335,
+      "grad_norm": 7.154104232788086,
+      "learning_rate": 6.6372e-08,
+      "loss": 0.4928,
+      "step": 233500
+    },
+    {
+      "epoch": 2.3355,
+      "grad_norm": 2.846604824066162,
+      "learning_rate": 6.617199999999999e-08,
+      "loss": 0.4661,
+      "step": 233550
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 93.85885620117188,
+      "learning_rate": 6.5972e-08,
+      "loss": 0.5282,
+      "step": 233600
+    },
+    {
+      "epoch": 2.3365,
+      "grad_norm": 14.930121421813965,
+      "learning_rate": 6.577199999999999e-08,
+      "loss": 0.3968,
+      "step": 233650
+    },
+    {
+      "epoch": 2.337,
+      "grad_norm": 50.37242889404297,
+      "learning_rate": 6.5572e-08,
+      "loss": 0.4821,
+      "step": 233700
+    },
+    {
+      "epoch": 2.3375,
+      "grad_norm": 7.079569339752197,
+      "learning_rate": 6.5372e-08,
+      "loss": 0.4573,
+      "step": 233750
+    },
+    {
+      "epoch": 2.338,
+      "grad_norm": 59.11067581176758,
+      "learning_rate": 6.5172e-08,
+      "loss": 0.4407,
+      "step": 233800
+    },
+    {
+      "epoch": 2.3385,
+      "grad_norm": 106.90413665771484,
+      "learning_rate": 6.4972e-08,
+      "loss": 0.5134,
+      "step": 233850
+    },
+    {
+      "epoch": 2.339,
+      "grad_norm": 88.19550323486328,
+      "learning_rate": 6.477199999999999e-08,
+      "loss": 0.3665,
+      "step": 233900
+    },
+    {
+      "epoch": 2.3395,
+      "grad_norm": 3.8314621448516846,
+      "learning_rate": 6.4572e-08,
+      "loss": 0.3177,
+      "step": 233950
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 78.19933319091797,
+      "learning_rate": 6.4372e-08,
+      "loss": 0.4219,
+      "step": 234000
+    },
+    {
+      "epoch": 2.3405,
+      "grad_norm": 62.778011322021484,
+      "learning_rate": 6.4172e-08,
+      "loss": 0.3997,
+      "step": 234050
+    },
+    {
+      "epoch": 2.341,
+      "grad_norm": 73.71574401855469,
+      "learning_rate": 6.3972e-08,
+      "loss": 0.3116,
+      "step": 234100
+    },
+    {
+      "epoch": 2.3415,
+      "grad_norm": 54.49970245361328,
+      "learning_rate": 6.3772e-08,
+      "loss": 0.4426,
+      "step": 234150
+    },
+    {
+      "epoch": 2.342,
+      "grad_norm": 63.628170013427734,
+      "learning_rate": 6.3572e-08,
+      "loss": 0.4592,
+      "step": 234200
+    },
+    {
+      "epoch": 2.3425000000000002,
+      "grad_norm": 69.60954284667969,
+      "learning_rate": 6.337199999999999e-08,
+      "loss": 0.3635,
+      "step": 234250
+    },
+    {
+      "epoch": 2.343,
+      "grad_norm": 52.40366744995117,
+      "learning_rate": 6.3172e-08,
+      "loss": 0.538,
+      "step": 234300
+    },
+    {
+      "epoch": 2.3435,
+      "grad_norm": 86.80823516845703,
+      "learning_rate": 6.2972e-08,
+      "loss": 0.3849,
+      "step": 234350
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 67.82777404785156,
+      "learning_rate": 6.277199999999999e-08,
+      "loss": 0.5053,
+      "step": 234400
+    },
+    {
+      "epoch": 2.3445,
+      "grad_norm": 63.045921325683594,
+      "learning_rate": 6.2572e-08,
+      "loss": 0.3757,
+      "step": 234450
+    },
+    {
+      "epoch": 2.3449999999999998,
+      "grad_norm": 64.31387329101562,
+      "learning_rate": 6.2372e-08,
+      "loss": 0.4009,
+      "step": 234500
+    },
+    {
+      "epoch": 2.3455,
+      "grad_norm": 43.08381652832031,
+      "learning_rate": 6.2172e-08,
+      "loss": 0.4228,
+      "step": 234550
+    },
+    {
+      "epoch": 2.346,
+      "grad_norm": 25.41963768005371,
+      "learning_rate": 6.1972e-08,
+      "loss": 0.3963,
+      "step": 234600
+    },
+    {
+      "epoch": 2.3465,
+      "grad_norm": 45.17002487182617,
+      "learning_rate": 6.177199999999999e-08,
+      "loss": 0.4244,
+      "step": 234650
+    },
+    {
+      "epoch": 2.347,
+      "grad_norm": 64.86937713623047,
+      "learning_rate": 6.1572e-08,
+      "loss": 0.433,
+      "step": 234700
+    },
+    {
+      "epoch": 2.3475,
+      "grad_norm": 122.25826263427734,
+      "learning_rate": 6.1372e-08,
+      "loss": 0.5875,
+      "step": 234750
+    },
+    {
+      "epoch": 2.348,
+      "grad_norm": 1.6496562957763672,
+      "learning_rate": 6.1172e-08,
+      "loss": 0.3615,
+      "step": 234800
+    },
+    {
+      "epoch": 2.3485,
+      "grad_norm": 34.062599182128906,
+      "learning_rate": 6.0972e-08,
+      "loss": 0.4072,
+      "step": 234850
+    },
+    {
+      "epoch": 2.349,
+      "grad_norm": 1.5793612003326416,
+      "learning_rate": 6.0772e-08,
+      "loss": 0.4464,
+      "step": 234900
+    },
+    {
+      "epoch": 2.3495,
+      "grad_norm": 17.111398696899414,
+      "learning_rate": 6.0572e-08,
+      "loss": 0.3973,
+      "step": 234950
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.124202728271484,
+      "learning_rate": 6.037199999999999e-08,
+      "loss": 0.4139,
+      "step": 235000
+    },
+    {
+      "epoch": 2.3505,
+      "grad_norm": 20.528778076171875,
+      "learning_rate": 6.0172e-08,
+      "loss": 0.4128,
+      "step": 235050
+    },
+    {
+      "epoch": 2.351,
+      "grad_norm": 9.12598991394043,
+      "learning_rate": 5.997199999999999e-08,
+      "loss": 0.429,
+      "step": 235100
+    },
+    {
+      "epoch": 2.3515,
+      "grad_norm": 89.71222686767578,
+      "learning_rate": 5.9772e-08,
+      "loss": 0.4243,
+      "step": 235150
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 8.675117492675781,
+      "learning_rate": 5.9576e-08,
+      "loss": 0.5467,
+      "step": 235200
+    },
+    {
+      "epoch": 2.3525,
+      "grad_norm": 82.99100494384766,
+      "learning_rate": 5.9375999999999995e-08,
+      "loss": 0.3422,
+      "step": 235250
+    },
+    {
+      "epoch": 2.3529999999999998,
+      "grad_norm": 77.30717468261719,
+      "learning_rate": 5.9176e-08,
+      "loss": 0.4311,
+      "step": 235300
+    },
+    {
+      "epoch": 2.3535,
+      "grad_norm": 58.811668395996094,
+      "learning_rate": 5.8976e-08,
+      "loss": 0.2901,
+      "step": 235350
+    },
+    {
+      "epoch": 2.354,
+      "grad_norm": 18.311758041381836,
+      "learning_rate": 5.8775999999999996e-08,
+      "loss": 0.602,
+      "step": 235400
+    },
+    {
+      "epoch": 2.3545,
+      "grad_norm": 54.06341552734375,
+      "learning_rate": 5.8576e-08,
+      "loss": 0.3226,
+      "step": 235450
+    },
+    {
+      "epoch": 2.355,
+      "grad_norm": 3.558945655822754,
+      "learning_rate": 5.837599999999999e-08,
+      "loss": 0.5484,
+      "step": 235500
+    },
+    {
+      "epoch": 2.3555,
+      "grad_norm": 40.88227081298828,
+      "learning_rate": 5.8176e-08,
+      "loss": 0.2963,
+      "step": 235550
+    },
+    {
+      "epoch": 2.356,
+      "grad_norm": 4.487430572509766,
+      "learning_rate": 5.7979999999999994e-08,
+      "loss": 0.4298,
+      "step": 235600
+    },
+    {
+      "epoch": 2.3565,
+      "grad_norm": 83.76229858398438,
+      "learning_rate": 5.778e-08,
+      "loss": 0.4383,
+      "step": 235650
+    },
+    {
+      "epoch": 2.357,
+      "grad_norm": 25.388702392578125,
+      "learning_rate": 5.758e-08,
+      "loss": 0.4707,
+      "step": 235700
+    },
+    {
+      "epoch": 2.3575,
+      "grad_norm": 21.608808517456055,
+      "learning_rate": 5.7379999999999995e-08,
+      "loss": 0.3404,
+      "step": 235750
+    },
+    {
+      "epoch": 2.358,
+      "grad_norm": 95.14769744873047,
+      "learning_rate": 5.718e-08,
+      "loss": 0.4012,
+      "step": 235800
+    },
+    {
+      "epoch": 2.3585,
+      "grad_norm": 29.87248420715332,
+      "learning_rate": 5.698e-08,
+      "loss": 0.6228,
+      "step": 235850
+    },
+    {
+      "epoch": 2.359,
+      "grad_norm": 119.35411834716797,
+      "learning_rate": 5.6779999999999996e-08,
+      "loss": 0.4535,
+      "step": 235900
+    },
+    {
+      "epoch": 2.3595,
+      "grad_norm": 22.905563354492188,
+      "learning_rate": 5.6579999999999994e-08,
+      "loss": 0.2687,
+      "step": 235950
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 112.05744934082031,
+      "learning_rate": 5.638e-08,
+      "loss": 0.3915,
+      "step": 236000
+    },
+    {
+      "epoch": 2.3605,
+      "grad_norm": 51.91557693481445,
+      "learning_rate": 5.618e-08,
+      "loss": 0.4123,
+      "step": 236050
+    },
+    {
+      "epoch": 2.3609999999999998,
+      "grad_norm": 35.96802520751953,
+      "learning_rate": 5.598e-08,
+      "loss": 0.3404,
+      "step": 236100
+    },
+    {
+      "epoch": 2.3615,
+      "grad_norm": 39.31023406982422,
+      "learning_rate": 5.578e-08,
+      "loss": 0.362,
+      "step": 236150
+    },
+    {
+      "epoch": 2.362,
+      "grad_norm": 100.35054016113281,
+      "learning_rate": 5.557999999999999e-08,
+      "loss": 0.5586,
+      "step": 236200
+    },
+    {
+      "epoch": 2.3625,
+      "grad_norm": 0.3581673502922058,
+      "learning_rate": 5.538e-08,
+      "loss": 0.3348,
+      "step": 236250
+    },
+    {
+      "epoch": 2.363,
+      "grad_norm": 7.026209831237793,
+      "learning_rate": 5.5179999999999995e-08,
+      "loss": 0.4421,
+      "step": 236300
+    },
+    {
+      "epoch": 2.3635,
+      "grad_norm": 1.0426843166351318,
+      "learning_rate": 5.498e-08,
+      "loss": 0.4958,
+      "step": 236350
+    },
+    {
+      "epoch": 2.364,
+      "grad_norm": 86.80310821533203,
+      "learning_rate": 5.478e-08,
+      "loss": 0.3704,
+      "step": 236400
+    },
+    {
+      "epoch": 2.3645,
+      "grad_norm": 99.3370590209961,
+      "learning_rate": 5.4579999999999996e-08,
+      "loss": 0.5521,
+      "step": 236450
+    },
+    {
+      "epoch": 2.365,
+      "grad_norm": 73.29484558105469,
+      "learning_rate": 5.4384e-08,
+      "loss": 0.4607,
+      "step": 236500
+    },
+    {
+      "epoch": 2.3655,
+      "grad_norm": 94.66316986083984,
+      "learning_rate": 5.4184e-08,
+      "loss": 0.4448,
+      "step": 236550
+    },
+    {
+      "epoch": 2.366,
+      "grad_norm": 4.878780364990234,
+      "learning_rate": 5.3983999999999996e-08,
+      "loss": 0.3542,
+      "step": 236600
+    },
+    {
+      "epoch": 2.3665,
+      "grad_norm": 38.552696228027344,
+      "learning_rate": 5.3783999999999994e-08,
+      "loss": 0.3363,
+      "step": 236650
+    },
+    {
+      "epoch": 2.367,
+      "grad_norm": 106.73291015625,
+      "learning_rate": 5.3584e-08,
+      "loss": 0.3413,
+      "step": 236700
+    },
+    {
+      "epoch": 2.3675,
+      "grad_norm": 42.5386848449707,
+      "learning_rate": 5.3384e-08,
+      "loss": 0.3877,
+      "step": 236750
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 53.35752487182617,
+      "learning_rate": 5.3184e-08,
+      "loss": 0.4686,
+      "step": 236800
+    },
+    {
+      "epoch": 2.3685,
+      "grad_norm": 7.68386173248291,
+      "learning_rate": 5.2984e-08,
+      "loss": 0.4841,
+      "step": 236850
+    },
+    {
+      "epoch": 2.3689999999999998,
+      "grad_norm": 121.2600326538086,
+      "learning_rate": 5.2784e-08,
+      "loss": 0.4746,
+      "step": 236900
+    },
+    {
+      "epoch": 2.3695,
+      "grad_norm": 25.649150848388672,
+      "learning_rate": 5.2583999999999996e-08,
+      "loss": 0.365,
+      "step": 236950
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.756162643432617,
+      "learning_rate": 5.2383999999999994e-08,
+      "loss": 0.3344,
+      "step": 237000
+    },
+    {
+      "epoch": 2.3705,
+      "grad_norm": 20.484275817871094,
+      "learning_rate": 5.2184e-08,
+      "loss": 0.5115,
+      "step": 237050
+    },
+    {
+      "epoch": 2.371,
+      "grad_norm": 111.54764556884766,
+      "learning_rate": 5.1984e-08,
+      "loss": 0.3719,
+      "step": 237100
+    },
+    {
+      "epoch": 2.3715,
+      "grad_norm": 2.243807315826416,
+      "learning_rate": 5.1783999999999996e-08,
+      "loss": 0.4457,
+      "step": 237150
+    },
+    {
+      "epoch": 2.372,
+      "grad_norm": 76.7511978149414,
+      "learning_rate": 5.1583999999999994e-08,
+      "loss": 0.3587,
+      "step": 237200
+    },
+    {
+      "epoch": 2.3725,
+      "grad_norm": 82.79983520507812,
+      "learning_rate": 5.1384e-08,
+      "loss": 0.4723,
+      "step": 237250
+    },
+    {
+      "epoch": 2.373,
+      "grad_norm": 76.18513488769531,
+      "learning_rate": 5.1184e-08,
+      "loss": 0.4799,
+      "step": 237300
+    },
+    {
+      "epoch": 2.3735,
+      "grad_norm": 94.5965805053711,
+      "learning_rate": 5.0984e-08,
+      "loss": 0.4241,
+      "step": 237350
+    },
+    {
+      "epoch": 2.374,
+      "grad_norm": 0.5150562524795532,
+      "learning_rate": 5.0784e-08,
+      "loss": 0.4691,
+      "step": 237400
+    },
+    {
+      "epoch": 2.3745,
+      "grad_norm": 3.979444742202759,
+      "learning_rate": 5.058399999999999e-08,
+      "loss": 0.3656,
+      "step": 237450
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 49.238975524902344,
+      "learning_rate": 5.0383999999999996e-08,
+      "loss": 0.4712,
+      "step": 237500
+    },
+    {
+      "epoch": 2.3755,
+      "grad_norm": 82.24160766601562,
+      "learning_rate": 5.0183999999999995e-08,
+      "loss": 0.3209,
+      "step": 237550
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 106.60427856445312,
+      "learning_rate": 4.9984e-08,
+      "loss": 0.5374,
+      "step": 237600
+    },
+    {
+      "epoch": 2.3765,
+      "grad_norm": 5.549964904785156,
+      "learning_rate": 4.9784e-08,
+      "loss": 0.3677,
+      "step": 237650
+    },
+    {
+      "epoch": 2.377,
+      "grad_norm": 106.1153564453125,
+      "learning_rate": 4.9584e-08,
+      "loss": 0.5558,
+      "step": 237700
+    },
+    {
+      "epoch": 2.3775,
+      "grad_norm": 6.455342769622803,
+      "learning_rate": 4.9383999999999994e-08,
+      "loss": 0.53,
+      "step": 237750
+    },
+    {
+      "epoch": 2.378,
+      "grad_norm": 28.48943519592285,
+      "learning_rate": 4.9184e-08,
+      "loss": 0.618,
+      "step": 237800
+    },
+    {
+      "epoch": 2.3785,
+      "grad_norm": 3.647404432296753,
+      "learning_rate": 4.8984e-08,
+      "loss": 0.5573,
+      "step": 237850
+    },
+    {
+      "epoch": 2.379,
+      "grad_norm": 3.0441486835479736,
+      "learning_rate": 4.8784e-08,
+      "loss": 0.3326,
+      "step": 237900
+    },
+    {
+      "epoch": 2.3795,
+      "grad_norm": 44.93407440185547,
+      "learning_rate": 4.8584e-08,
+      "loss": 0.3113,
+      "step": 237950
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 73.76074981689453,
+      "learning_rate": 4.8384e-08,
+      "loss": 0.3362,
+      "step": 238000
+    },
+    {
+      "epoch": 2.3805,
+      "grad_norm": 7.347684383392334,
+      "learning_rate": 4.8183999999999997e-08,
+      "loss": 0.4527,
+      "step": 238050
+    },
+    {
+      "epoch": 2.3810000000000002,
+      "grad_norm": 61.60081481933594,
+      "learning_rate": 4.7983999999999995e-08,
+      "loss": 0.464,
+      "step": 238100
+    },
+    {
+      "epoch": 2.3815,
+      "grad_norm": 11.561792373657227,
+      "learning_rate": 4.7784e-08,
+      "loss": 0.5532,
+      "step": 238150
+    },
+    {
+      "epoch": 2.382,
+      "grad_norm": 1.321579933166504,
+      "learning_rate": 4.7584e-08,
+      "loss": 0.4313,
+      "step": 238200
+    },
+    {
+      "epoch": 2.3825,
+      "grad_norm": 35.636898040771484,
+      "learning_rate": 4.7384e-08,
+      "loss": 0.4452,
+      "step": 238250
+    },
+    {
+      "epoch": 2.383,
+      "grad_norm": 82.60275268554688,
+      "learning_rate": 4.7183999999999994e-08,
+      "loss": 0.4888,
+      "step": 238300
+    },
+    {
+      "epoch": 2.3835,
+      "grad_norm": 10.286293029785156,
+      "learning_rate": 4.6984e-08,
+      "loss": 0.328,
+      "step": 238350
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 126.57600402832031,
+      "learning_rate": 4.6784e-08,
+      "loss": 0.4364,
+      "step": 238400
+    },
+    {
+      "epoch": 2.3845,
+      "grad_norm": 0.09401095658540726,
+      "learning_rate": 4.6583999999999995e-08,
+      "loss": 0.3141,
+      "step": 238450
+    },
+    {
+      "epoch": 2.385,
+      "grad_norm": 42.13533401489258,
+      "learning_rate": 4.6384e-08,
+      "loss": 0.5262,
+      "step": 238500
+    },
+    {
+      "epoch": 2.3855,
+      "grad_norm": 8.396677017211914,
+      "learning_rate": 4.6184e-08,
+      "loss": 0.3766,
+      "step": 238550
+    },
+    {
+      "epoch": 2.386,
+      "grad_norm": 12.288100242614746,
+      "learning_rate": 4.5984e-08,
+      "loss": 0.4144,
+      "step": 238600
+    },
+    {
+      "epoch": 2.3865,
+      "grad_norm": 81.89070892333984,
+      "learning_rate": 4.5783999999999995e-08,
+      "loss": 0.3493,
+      "step": 238650
+    },
+    {
+      "epoch": 2.387,
+      "grad_norm": 104.77812194824219,
+      "learning_rate": 4.5584e-08,
+      "loss": 0.329,
+      "step": 238700
+    },
+    {
+      "epoch": 2.3875,
+      "grad_norm": 111.44982147216797,
+      "learning_rate": 4.5384e-08,
+      "loss": 0.3988,
+      "step": 238750
+    },
+    {
+      "epoch": 2.388,
+      "grad_norm": 96.20793151855469,
+      "learning_rate": 4.5184e-08,
+      "loss": 0.2894,
+      "step": 238800
+    },
+    {
+      "epoch": 2.3885,
+      "grad_norm": 67.85469818115234,
+      "learning_rate": 4.4984e-08,
+      "loss": 0.3836,
+      "step": 238850
+    },
+    {
+      "epoch": 2.3890000000000002,
+      "grad_norm": 53.94377899169922,
+      "learning_rate": 4.478399999999999e-08,
+      "loss": 0.4408,
+      "step": 238900
+    },
+    {
+      "epoch": 2.3895,
+      "grad_norm": 174.99449157714844,
+      "learning_rate": 4.4584e-08,
+      "loss": 0.454,
+      "step": 238950
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.08309555053711,
+      "learning_rate": 4.4383999999999996e-08,
+      "loss": 0.4012,
+      "step": 239000
+    },
+    {
+      "epoch": 2.3905,
+      "grad_norm": 7.337883949279785,
+      "learning_rate": 4.4184e-08,
+      "loss": 0.4567,
+      "step": 239050
+    },
+    {
+      "epoch": 2.391,
+      "grad_norm": 118.68801879882812,
+      "learning_rate": 4.3984e-08,
+      "loss": 0.4913,
+      "step": 239100
+    },
+    {
+      "epoch": 2.3915,
+      "grad_norm": 119.5619888305664,
+      "learning_rate": 4.3784e-08,
+      "loss": 0.3864,
+      "step": 239150
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 63.690818786621094,
+      "learning_rate": 4.3583999999999995e-08,
+      "loss": 0.3754,
+      "step": 239200
+    },
+    {
+      "epoch": 2.3925,
+      "grad_norm": 15.803923606872559,
+      "learning_rate": 4.3384e-08,
+      "loss": 0.4492,
+      "step": 239250
+    },
+    {
+      "epoch": 2.393,
+      "grad_norm": 13.975971221923828,
+      "learning_rate": 4.3184e-08,
+      "loss": 0.3106,
+      "step": 239300
+    },
+    {
+      "epoch": 2.3935,
+      "grad_norm": 6.938758373260498,
+      "learning_rate": 4.2983999999999996e-08,
+      "loss": 0.3647,
+      "step": 239350
+    },
+    {
+      "epoch": 2.394,
+      "grad_norm": 80.98222351074219,
+      "learning_rate": 4.2784e-08,
+      "loss": 0.365,
+      "step": 239400
+    },
+    {
+      "epoch": 2.3945,
+      "grad_norm": 11.557998657226562,
+      "learning_rate": 4.258399999999999e-08,
+      "loss": 0.4769,
+      "step": 239450
+    },
+    {
+      "epoch": 2.395,
+      "grad_norm": 20.222549438476562,
+      "learning_rate": 4.2384e-08,
+      "loss": 0.5143,
+      "step": 239500
+    },
+    {
+      "epoch": 2.3955,
+      "grad_norm": 61.61232376098633,
+      "learning_rate": 4.2183999999999996e-08,
+      "loss": 0.4041,
+      "step": 239550
+    },
+    {
+      "epoch": 2.396,
+      "grad_norm": 40.21515655517578,
+      "learning_rate": 4.1988e-08,
+      "loss": 0.4189,
+      "step": 239600
+    },
+    {
+      "epoch": 2.3965,
+      "grad_norm": 39.18798065185547,
+      "learning_rate": 4.1788e-08,
+      "loss": 0.2413,
+      "step": 239650
+    },
+    {
+      "epoch": 2.3970000000000002,
+      "grad_norm": 0.4177449345588684,
+      "learning_rate": 4.1587999999999995e-08,
+      "loss": 0.3866,
+      "step": 239700
+    },
+    {
+      "epoch": 2.3975,
+      "grad_norm": 32.55744171142578,
+      "learning_rate": 4.1388e-08,
+      "loss": 0.3252,
+      "step": 239750
+    },
+    {
+      "epoch": 2.398,
+      "grad_norm": 93.50446319580078,
+      "learning_rate": 4.1188e-08,
+      "loss": 0.5187,
+      "step": 239800
+    },
+    {
+      "epoch": 2.3985,
+      "grad_norm": 21.071910858154297,
+      "learning_rate": 4.0987999999999996e-08,
+      "loss": 0.4274,
+      "step": 239850
+    },
+    {
+      "epoch": 2.399,
+      "grad_norm": 156.43629455566406,
+      "learning_rate": 4.0787999999999994e-08,
+      "loss": 0.4526,
+      "step": 239900
+    },
+    {
+      "epoch": 2.3995,
+      "grad_norm": 71.36151123046875,
+      "learning_rate": 4.0588e-08,
+      "loss": 0.4707,
+      "step": 239950
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.978219985961914,
+      "learning_rate": 4.0388e-08,
+      "loss": 0.4547,
+      "step": 240000
+    },
+    {
+      "epoch": 2.4005,
+      "grad_norm": 17.83655548095703,
+      "learning_rate": 4.0188e-08,
+      "loss": 0.5657,
+      "step": 240050
+    },
+    {
+      "epoch": 2.401,
+      "grad_norm": 139.66326904296875,
+      "learning_rate": 3.9988e-08,
+      "loss": 0.356,
+      "step": 240100
+    },
+    {
+      "epoch": 2.4015,
+      "grad_norm": 20.223098754882812,
+      "learning_rate": 3.978799999999999e-08,
+      "loss": 0.382,
+      "step": 240150
+    },
+    {
+      "epoch": 2.402,
+      "grad_norm": 2.374896287918091,
+      "learning_rate": 3.9588e-08,
+      "loss": 0.4235,
+      "step": 240200
+    },
+    {
+      "epoch": 2.4025,
+      "grad_norm": 60.51106643676758,
+      "learning_rate": 3.9387999999999995e-08,
+      "loss": 0.6068,
+      "step": 240250
+    },
+    {
+      "epoch": 2.403,
+      "grad_norm": 72.06840515136719,
+      "learning_rate": 3.9188e-08,
+      "loss": 0.474,
+      "step": 240300
+    },
+    {
+      "epoch": 2.4035,
+      "grad_norm": 78.26203155517578,
+      "learning_rate": 3.8988e-08,
+      "loss": 0.4211,
+      "step": 240350
+    },
+    {
+      "epoch": 2.404,
+      "grad_norm": 30.761234283447266,
+      "learning_rate": 3.8788e-08,
+      "loss": 0.2871,
+      "step": 240400
+    },
+    {
+      "epoch": 2.4045,
+      "grad_norm": 20.948881149291992,
+      "learning_rate": 3.8587999999999995e-08,
+      "loss": 0.4948,
+      "step": 240450
+    },
+    {
+      "epoch": 2.4050000000000002,
+      "grad_norm": 2.7093822956085205,
+      "learning_rate": 3.8388e-08,
+      "loss": 0.256,
+      "step": 240500
+    },
+    {
+      "epoch": 2.4055,
+      "grad_norm": 8.741806983947754,
+      "learning_rate": 3.8188e-08,
+      "loss": 0.4368,
+      "step": 240550
+    },
+    {
+      "epoch": 2.406,
+      "grad_norm": 13.452826499938965,
+      "learning_rate": 3.7988e-08,
+      "loss": 0.3249,
+      "step": 240600
+    },
+    {
+      "epoch": 2.4065,
+      "grad_norm": 9.17270565032959,
+      "learning_rate": 3.7788e-08,
+      "loss": 0.434,
+      "step": 240650
+    },
+    {
+      "epoch": 2.407,
+      "grad_norm": 44.4083366394043,
+      "learning_rate": 3.7588e-08,
+      "loss": 0.37,
+      "step": 240700
+    },
+    {
+      "epoch": 2.4074999999999998,
+      "grad_norm": 5.229351997375488,
+      "learning_rate": 3.7388e-08,
+      "loss": 0.3604,
+      "step": 240750
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 113.33466339111328,
+      "learning_rate": 3.7187999999999995e-08,
+      "loss": 0.3738,
+      "step": 240800
+    },
+    {
+      "epoch": 2.4085,
+      "grad_norm": 66.20381927490234,
+      "learning_rate": 3.6988e-08,
+      "loss": 0.3353,
+      "step": 240850
+    },
+    {
+      "epoch": 2.409,
+      "grad_norm": 15.30612850189209,
+      "learning_rate": 3.6788e-08,
+      "loss": 0.2977,
+      "step": 240900
+    },
+    {
+      "epoch": 2.4095,
+      "grad_norm": 0.48715662956237793,
+      "learning_rate": 3.6588e-08,
+      "loss": 0.3838,
+      "step": 240950
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 21.8796329498291,
+      "learning_rate": 3.6387999999999995e-08,
+      "loss": 0.5768,
+      "step": 241000
+    },
+    {
+      "epoch": 2.4105,
+      "grad_norm": 58.25382995605469,
+      "learning_rate": 3.6188e-08,
+      "loss": 0.222,
+      "step": 241050
+    },
+    {
+      "epoch": 2.411,
+      "grad_norm": 75.87273406982422,
+      "learning_rate": 3.5988e-08,
+      "loss": 0.3897,
+      "step": 241100
+    },
+    {
+      "epoch": 2.4115,
+      "grad_norm": 132.18612670898438,
+      "learning_rate": 3.5787999999999996e-08,
+      "loss": 0.4991,
+      "step": 241150
+    },
+    {
+      "epoch": 2.412,
+      "grad_norm": 93.72396087646484,
+      "learning_rate": 3.5588e-08,
+      "loss": 0.5212,
+      "step": 241200
+    },
+    {
+      "epoch": 2.4125,
+      "grad_norm": 73.5746841430664,
+      "learning_rate": 3.5388e-08,
+      "loss": 0.4832,
+      "step": 241250
+    },
+    {
+      "epoch": 2.413,
+      "grad_norm": 0.44722920656204224,
+      "learning_rate": 3.5188e-08,
+      "loss": 0.3307,
+      "step": 241300
+    },
+    {
+      "epoch": 2.4135,
+      "grad_norm": 12.440936088562012,
+      "learning_rate": 3.4987999999999995e-08,
+      "loss": 0.4114,
+      "step": 241350
+    },
+    {
+      "epoch": 2.414,
+      "grad_norm": 5.505441665649414,
+      "learning_rate": 3.4788e-08,
+      "loss": 0.4136,
+      "step": 241400
+    },
+    {
+      "epoch": 2.4145,
+      "grad_norm": 54.741905212402344,
+      "learning_rate": 3.4588e-08,
+      "loss": 0.4845,
+      "step": 241450
+    },
+    {
+      "epoch": 2.415,
+      "grad_norm": 3.8523218631744385,
+      "learning_rate": 3.4388e-08,
+      "loss": 0.3512,
+      "step": 241500
+    },
+    {
+      "epoch": 2.4154999999999998,
+      "grad_norm": 60.24558639526367,
+      "learning_rate": 3.4188e-08,
+      "loss": 0.43,
+      "step": 241550
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 103.20320129394531,
+      "learning_rate": 3.398799999999999e-08,
+      "loss": 0.5293,
+      "step": 241600
+    },
+    {
+      "epoch": 2.4165,
+      "grad_norm": 43.67532730102539,
+      "learning_rate": 3.3788e-08,
+      "loss": 0.4799,
+      "step": 241650
+    },
+    {
+      "epoch": 2.417,
+      "grad_norm": 12.198750495910645,
+      "learning_rate": 3.3587999999999996e-08,
+      "loss": 0.4764,
+      "step": 241700
+    },
+    {
+      "epoch": 2.4175,
+      "grad_norm": 44.76414489746094,
+      "learning_rate": 3.3388e-08,
+      "loss": 0.3694,
+      "step": 241750
+    },
+    {
+      "epoch": 2.418,
+      "grad_norm": 93.81781005859375,
+      "learning_rate": 3.3188e-08,
+      "loss": 0.4577,
+      "step": 241800
+    },
+    {
+      "epoch": 2.4185,
+      "grad_norm": 86.02700805664062,
+      "learning_rate": 3.2988000000000004e-08,
+      "loss": 0.337,
+      "step": 241850
+    },
+    {
+      "epoch": 2.419,
+      "grad_norm": 23.790729522705078,
+      "learning_rate": 3.2787999999999996e-08,
+      "loss": 0.3581,
+      "step": 241900
+    },
+    {
+      "epoch": 2.4195,
+      "grad_norm": 80.7553482055664,
+      "learning_rate": 3.2588e-08,
+      "loss": 0.4819,
+      "step": 241950
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 100.54222106933594,
+      "learning_rate": 3.2388e-08,
+      "loss": 0.3009,
+      "step": 242000
+    },
+    {
+      "epoch": 2.4205,
+      "grad_norm": 69.31678009033203,
+      "learning_rate": 3.2188e-08,
+      "loss": 0.414,
+      "step": 242050
+    },
+    {
+      "epoch": 2.421,
+      "grad_norm": 8.001131057739258,
+      "learning_rate": 3.1988e-08,
+      "loss": 0.5114,
+      "step": 242100
+    },
+    {
+      "epoch": 2.4215,
+      "grad_norm": 96.0322036743164,
+      "learning_rate": 3.178799999999999e-08,
+      "loss": 0.4587,
+      "step": 242150
+    },
+    {
+      "epoch": 2.422,
+      "grad_norm": 8.513936042785645,
+      "learning_rate": 3.1588e-08,
+      "loss": 0.3816,
+      "step": 242200
+    },
+    {
+      "epoch": 2.4225,
+      "grad_norm": 53.89067077636719,
+      "learning_rate": 3.1387999999999996e-08,
+      "loss": 0.4928,
+      "step": 242250
+    },
+    {
+      "epoch": 2.423,
+      "grad_norm": 98.1548080444336,
+      "learning_rate": 3.1188e-08,
+      "loss": 0.4295,
+      "step": 242300
+    },
+    {
+      "epoch": 2.4234999999999998,
+      "grad_norm": 4.35904598236084,
+      "learning_rate": 3.0988e-08,
+      "loss": 0.4473,
+      "step": 242350
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 72.66747283935547,
+      "learning_rate": 3.0788e-08,
+      "loss": 0.4041,
+      "step": 242400
+    },
+    {
+      "epoch": 2.4245,
+      "grad_norm": 275.8028259277344,
+      "learning_rate": 3.0588e-08,
+      "loss": 0.384,
+      "step": 242450
+    },
+    {
+      "epoch": 2.425,
+      "grad_norm": 74.81790924072266,
+      "learning_rate": 3.0387999999999994e-08,
+      "loss": 0.4163,
+      "step": 242500
+    },
+    {
+      "epoch": 2.4255,
+      "grad_norm": 2.4196512699127197,
+      "learning_rate": 3.0188e-08,
+      "loss": 0.4067,
+      "step": 242550
+    },
+    {
+      "epoch": 2.426,
+      "grad_norm": 60.739994049072266,
+      "learning_rate": 2.9988e-08,
+      "loss": 0.4061,
+      "step": 242600
+    },
+    {
+      "epoch": 2.4265,
+      "grad_norm": 11.086175918579102,
+      "learning_rate": 2.9788e-08,
+      "loss": 0.343,
+      "step": 242650
+    },
+    {
+      "epoch": 2.427,
+      "grad_norm": 9.039807319641113,
+      "learning_rate": 2.9588e-08,
+      "loss": 0.3752,
+      "step": 242700
+    },
+    {
+      "epoch": 2.4275,
+      "grad_norm": 1.2201632261276245,
+      "learning_rate": 2.9387999999999998e-08,
+      "loss": 0.3769,
+      "step": 242750
+    },
+    {
+      "epoch": 2.428,
+      "grad_norm": 46.109474182128906,
+      "learning_rate": 2.9187999999999996e-08,
+      "loss": 0.4977,
+      "step": 242800
+    },
+    {
+      "epoch": 2.4285,
+      "grad_norm": 4.451110363006592,
+      "learning_rate": 2.8987999999999998e-08,
+      "loss": 0.4644,
+      "step": 242850
+    },
+    {
+      "epoch": 2.429,
+      "grad_norm": 78.92597961425781,
+      "learning_rate": 2.8788e-08,
+      "loss": 0.3742,
+      "step": 242900
+    },
+    {
+      "epoch": 2.4295,
+      "grad_norm": 4.9124908447265625,
+      "learning_rate": 2.8587999999999998e-08,
+      "loss": 0.3717,
+      "step": 242950
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.718141555786133,
+      "learning_rate": 2.8388e-08,
+      "loss": 0.4299,
+      "step": 243000
+    },
+    {
+      "epoch": 2.4305,
+      "grad_norm": 95.63214874267578,
+      "learning_rate": 2.8188e-08,
+      "loss": 0.3924,
+      "step": 243050
+    },
+    {
+      "epoch": 2.431,
+      "grad_norm": 7.528963088989258,
+      "learning_rate": 2.7988e-08,
+      "loss": 0.5847,
+      "step": 243100
+    },
+    {
+      "epoch": 2.4314999999999998,
+      "grad_norm": 2.7887654304504395,
+      "learning_rate": 2.7788e-08,
+      "loss": 0.3462,
+      "step": 243150
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 17.53578758239746,
+      "learning_rate": 2.7588e-08,
+      "loss": 0.4067,
+      "step": 243200
+    },
+    {
+      "epoch": 2.4325,
+      "grad_norm": 88.1779556274414,
+      "learning_rate": 2.7387999999999997e-08,
+      "loss": 0.3385,
+      "step": 243250
+    },
+    {
+      "epoch": 2.433,
+      "grad_norm": 100.06237030029297,
+      "learning_rate": 2.7187999999999998e-08,
+      "loss": 0.4189,
+      "step": 243300
+    },
+    {
+      "epoch": 2.4335,
+      "grad_norm": 89.0728759765625,
+      "learning_rate": 2.6988e-08,
+      "loss": 0.3789,
+      "step": 243350
+    },
+    {
+      "epoch": 2.434,
+      "grad_norm": 0.17192931473255157,
+      "learning_rate": 2.6787999999999998e-08,
+      "loss": 0.3695,
+      "step": 243400
+    },
+    {
+      "epoch": 2.4345,
+      "grad_norm": 8.71643352508545,
+      "learning_rate": 2.6588e-08,
+      "loss": 0.5545,
+      "step": 243450
+    },
+    {
+      "epoch": 2.435,
+      "grad_norm": 14.25753116607666,
+      "learning_rate": 2.6388e-08,
+      "loss": 0.4481,
+      "step": 243500
+    },
+    {
+      "epoch": 2.4355,
+      "grad_norm": 29.616668701171875,
+      "learning_rate": 2.6188e-08,
+      "loss": 0.2906,
+      "step": 243550
+    },
+    {
+      "epoch": 2.436,
+      "grad_norm": 126.89151000976562,
+      "learning_rate": 2.5988e-08,
+      "loss": 0.5021,
+      "step": 243600
+    },
+    {
+      "epoch": 2.4365,
+      "grad_norm": 36.65034103393555,
+      "learning_rate": 2.5787999999999996e-08,
+      "loss": 0.4752,
+      "step": 243650
+    },
+    {
+      "epoch": 2.437,
+      "grad_norm": 52.342201232910156,
+      "learning_rate": 2.5587999999999997e-08,
+      "loss": 0.4993,
+      "step": 243700
+    },
+    {
+      "epoch": 2.4375,
+      "grad_norm": 14.278721809387207,
+      "learning_rate": 2.5388e-08,
+      "loss": 0.4432,
+      "step": 243750
+    },
+    {
+      "epoch": 2.438,
+      "grad_norm": 100.52146911621094,
+      "learning_rate": 2.5187999999999997e-08,
+      "loss": 0.4158,
+      "step": 243800
+    },
+    {
+      "epoch": 2.4385,
+      "grad_norm": 1.8151702880859375,
+      "learning_rate": 2.4988e-08,
+      "loss": 0.3256,
+      "step": 243850
+    },
+    {
+      "epoch": 2.439,
+      "grad_norm": 2.068545341491699,
+      "learning_rate": 2.4788e-08,
+      "loss": 0.3985,
+      "step": 243900
+    },
+    {
+      "epoch": 2.4395,
+      "grad_norm": 114.05219268798828,
+      "learning_rate": 2.4587999999999998e-08,
+      "loss": 0.5039,
+      "step": 243950
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.46074104309082,
+      "learning_rate": 2.4388e-08,
+      "loss": 0.3068,
+      "step": 244000
+    },
+    {
+      "epoch": 2.4405,
+      "grad_norm": 49.89002990722656,
+      "learning_rate": 2.4188e-08,
+      "loss": 0.5444,
+      "step": 244050
+    },
+    {
+      "epoch": 2.441,
+      "grad_norm": 116.9295883178711,
+      "learning_rate": 2.3988e-08,
+      "loss": 0.401,
+      "step": 244100
+    },
+    {
+      "epoch": 2.4415,
+      "grad_norm": 159.10597229003906,
+      "learning_rate": 2.3787999999999998e-08,
+      "loss": 0.5854,
+      "step": 244150
+    },
+    {
+      "epoch": 2.442,
+      "grad_norm": 72.9452896118164,
+      "learning_rate": 2.3588e-08,
+      "loss": 0.5169,
+      "step": 244200
+    },
+    {
+      "epoch": 2.4425,
+      "grad_norm": 59.4945068359375,
+      "learning_rate": 2.3387999999999997e-08,
+      "loss": 0.5033,
+      "step": 244250
+    },
+    {
+      "epoch": 2.443,
+      "grad_norm": 110.84677124023438,
+      "learning_rate": 2.3188e-08,
+      "loss": 0.446,
+      "step": 244300
+    },
+    {
+      "epoch": 2.4435000000000002,
+      "grad_norm": 46.814674377441406,
+      "learning_rate": 2.2988e-08,
+      "loss": 0.4448,
+      "step": 244350
+    },
+    {
+      "epoch": 2.444,
+      "grad_norm": 42.03607940673828,
+      "learning_rate": 2.2788e-08,
+      "loss": 0.4175,
+      "step": 244400
+    },
+    {
+      "epoch": 2.4445,
+      "grad_norm": 4.896573543548584,
+      "learning_rate": 2.2588e-08,
+      "loss": 0.3549,
+      "step": 244450
+    },
+    {
+      "epoch": 2.445,
+      "grad_norm": 66.42455291748047,
+      "learning_rate": 2.2388000000000002e-08,
+      "loss": 0.5136,
+      "step": 244500
+    },
+    {
+      "epoch": 2.4455,
+      "grad_norm": 9.06808090209961,
+      "learning_rate": 2.2188e-08,
+      "loss": 0.4,
+      "step": 244550
+    },
+    {
+      "epoch": 2.446,
+      "grad_norm": 32.235755920410156,
+      "learning_rate": 2.1987999999999998e-08,
+      "loss": 0.3744,
+      "step": 244600
+    },
+    {
+      "epoch": 2.4465,
+      "grad_norm": 87.67384338378906,
+      "learning_rate": 2.1787999999999996e-08,
+      "loss": 0.4273,
+      "step": 244650
+    },
+    {
+      "epoch": 2.447,
+      "grad_norm": 72.4257583618164,
+      "learning_rate": 2.1587999999999998e-08,
+      "loss": 0.5571,
+      "step": 244700
+    },
+    {
+      "epoch": 2.4475,
+      "grad_norm": 0.4879385530948639,
+      "learning_rate": 2.1388e-08,
+      "loss": 0.5196,
+      "step": 244750
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 80.70654296875,
+      "learning_rate": 2.1187999999999998e-08,
+      "loss": 0.5049,
+      "step": 244800
+    },
+    {
+      "epoch": 2.4485,
+      "grad_norm": 38.23407745361328,
+      "learning_rate": 2.0988e-08,
+      "loss": 0.3874,
+      "step": 244850
+    },
+    {
+      "epoch": 2.449,
+      "grad_norm": 31.929292678833008,
+      "learning_rate": 2.0788e-08,
+      "loss": 0.433,
+      "step": 244900
+    },
+    {
+      "epoch": 2.4495,
+      "grad_norm": 21.76283836364746,
+      "learning_rate": 2.0588e-08,
+      "loss": 0.41,
+      "step": 244950
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.751602172851562,
+      "learning_rate": 2.0391999999999998e-08,
+      "loss": 0.3788,
+      "step": 245000
+    },
+    {
+      "epoch": 2.4505,
+      "grad_norm": 10.776887893676758,
+      "learning_rate": 2.0192e-08,
+      "loss": 0.4291,
+      "step": 245050
+    },
+    {
+      "epoch": 2.451,
+      "grad_norm": 0.36848610639572144,
+      "learning_rate": 1.9991999999999998e-08,
+      "loss": 0.4365,
+      "step": 245100
+    },
+    {
+      "epoch": 2.4515000000000002,
+      "grad_norm": 23.090919494628906,
+      "learning_rate": 1.9792e-08,
+      "loss": 0.406,
+      "step": 245150
+    },
+    {
+      "epoch": 2.452,
+      "grad_norm": 25.339353561401367,
+      "learning_rate": 1.9591999999999998e-08,
+      "loss": 0.4708,
+      "step": 245200
+    },
+    {
+      "epoch": 2.4525,
+      "grad_norm": 52.52699661254883,
+      "learning_rate": 1.9392e-08,
+      "loss": 0.4524,
+      "step": 245250
+    },
+    {
+      "epoch": 2.453,
+      "grad_norm": 1.461986780166626,
+      "learning_rate": 1.9192e-08,
+      "loss": 0.4664,
+      "step": 245300
+    },
+    {
+      "epoch": 2.4535,
+      "grad_norm": 95.0424575805664,
+      "learning_rate": 1.8992e-08,
+      "loss": 0.3294,
+      "step": 245350
+    },
+    {
+      "epoch": 2.454,
+      "grad_norm": 91.5193862915039,
+      "learning_rate": 1.8792e-08,
+      "loss": 0.3741,
+      "step": 245400
+    },
+    {
+      "epoch": 2.4545,
+      "grad_norm": 95.59048461914062,
+      "learning_rate": 1.8592e-08,
+      "loss": 0.5858,
+      "step": 245450
+    },
+    {
+      "epoch": 2.455,
+      "grad_norm": 1.0647406578063965,
+      "learning_rate": 1.8391999999999997e-08,
+      "loss": 0.3868,
+      "step": 245500
+    },
+    {
+      "epoch": 2.4555,
+      "grad_norm": 36.12507629394531,
+      "learning_rate": 1.8192e-08,
+      "loss": 0.269,
+      "step": 245550
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 31.336393356323242,
+      "learning_rate": 1.7992e-08,
+      "loss": 0.388,
+      "step": 245600
+    },
+    {
+      "epoch": 2.4565,
+      "grad_norm": 103.9297866821289,
+      "learning_rate": 1.7791999999999998e-08,
+      "loss": 0.4473,
+      "step": 245650
+    },
+    {
+      "epoch": 2.457,
+      "grad_norm": 63.19395446777344,
+      "learning_rate": 1.7592e-08,
+      "loss": 0.4356,
+      "step": 245700
+    },
+    {
+      "epoch": 2.4575,
+      "grad_norm": 72.82255554199219,
+      "learning_rate": 1.7392e-08,
+      "loss": 0.434,
+      "step": 245750
+    },
+    {
+      "epoch": 2.458,
+      "grad_norm": 25.266742706298828,
+      "learning_rate": 1.7192e-08,
+      "loss": 0.4078,
+      "step": 245800
+    },
+    {
+      "epoch": 2.4585,
+      "grad_norm": 26.895408630371094,
+      "learning_rate": 1.6992e-08,
+      "loss": 0.5463,
+      "step": 245850
+    },
+    {
+      "epoch": 2.459,
+      "grad_norm": 18.793058395385742,
+      "learning_rate": 1.6792e-08,
+      "loss": 0.4784,
+      "step": 245900
+    },
+    {
+      "epoch": 2.4595000000000002,
+      "grad_norm": 1.1856598854064941,
+      "learning_rate": 1.6591999999999997e-08,
+      "loss": 0.3484,
+      "step": 245950
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.1075769662857056,
+      "learning_rate": 1.6392e-08,
+      "loss": 0.4504,
+      "step": 246000
+    },
+    {
+      "epoch": 2.4605,
+      "grad_norm": 47.581871032714844,
+      "learning_rate": 1.6192e-08,
+      "loss": 0.4725,
+      "step": 246050
+    },
+    {
+      "epoch": 2.461,
+      "grad_norm": 132.03213500976562,
+      "learning_rate": 1.5992e-08,
+      "loss": 0.46,
+      "step": 246100
+    },
+    {
+      "epoch": 2.4615,
+      "grad_norm": 129.3053741455078,
+      "learning_rate": 1.5792e-08,
+      "loss": 0.4021,
+      "step": 246150
+    },
+    {
+      "epoch": 2.462,
+      "grad_norm": 94.52012634277344,
+      "learning_rate": 1.5591999999999998e-08,
+      "loss": 0.5046,
+      "step": 246200
+    },
+    {
+      "epoch": 2.4625,
+      "grad_norm": 2.470003128051758,
+      "learning_rate": 1.5392e-08,
+      "loss": 0.4378,
+      "step": 246250
+    },
+    {
+      "epoch": 2.463,
+      "grad_norm": 65.46109008789062,
+      "learning_rate": 1.5192e-08,
+      "loss": 0.3905,
+      "step": 246300
+    },
+    {
+      "epoch": 2.4635,
+      "grad_norm": 29.295551300048828,
+      "learning_rate": 1.4992e-08,
+      "loss": 0.4354,
+      "step": 246350
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 46.023048400878906,
+      "learning_rate": 1.4792e-08,
+      "loss": 0.3686,
+      "step": 246400
+    },
+    {
+      "epoch": 2.4645,
+      "grad_norm": 2.7170820236206055,
+      "learning_rate": 1.4592000000000001e-08,
+      "loss": 0.4175,
+      "step": 246450
+    },
+    {
+      "epoch": 2.465,
+      "grad_norm": 46.63579559326172,
+      "learning_rate": 1.4391999999999999e-08,
+      "loss": 0.476,
+      "step": 246500
+    },
+    {
+      "epoch": 2.4655,
+      "grad_norm": 31.72504997253418,
+      "learning_rate": 1.4191999999999999e-08,
+      "loss": 0.4872,
+      "step": 246550
+    },
+    {
+      "epoch": 2.466,
+      "grad_norm": 2.155219793319702,
+      "learning_rate": 1.3991999999999999e-08,
+      "loss": 0.3661,
+      "step": 246600
+    },
+    {
+      "epoch": 2.4665,
+      "grad_norm": 100.10980987548828,
+      "learning_rate": 1.3792e-08,
+      "loss": 0.4025,
+      "step": 246650
+    },
+    {
+      "epoch": 2.467,
+      "grad_norm": 40.943275451660156,
+      "learning_rate": 1.3591999999999999e-08,
+      "loss": 0.4854,
+      "step": 246700
+    },
+    {
+      "epoch": 2.4675000000000002,
+      "grad_norm": 68.8076400756836,
+      "learning_rate": 1.3391999999999998e-08,
+      "loss": 0.3888,
+      "step": 246750
+    },
+    {
+      "epoch": 2.468,
+      "grad_norm": 53.90141296386719,
+      "learning_rate": 1.3192e-08,
+      "loss": 0.3383,
+      "step": 246800
+    },
+    {
+      "epoch": 2.4685,
+      "grad_norm": 65.13260650634766,
+      "learning_rate": 1.2992e-08,
+      "loss": 0.4943,
+      "step": 246850
+    },
+    {
+      "epoch": 2.469,
+      "grad_norm": 9.727241516113281,
+      "learning_rate": 1.2792e-08,
+      "loss": 0.5687,
+      "step": 246900
+    },
+    {
+      "epoch": 2.4695,
+      "grad_norm": 45.80133056640625,
+      "learning_rate": 1.2592e-08,
+      "loss": 0.2708,
+      "step": 246950
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 95.85955810546875,
+      "learning_rate": 1.2392e-08,
+      "loss": 0.4595,
+      "step": 247000
+    },
+    {
+      "epoch": 2.4705,
+      "grad_norm": 1.5538358688354492,
+      "learning_rate": 1.2192e-08,
+      "loss": 0.2399,
+      "step": 247050
+    },
+    {
+      "epoch": 2.471,
+      "grad_norm": 87.2671890258789,
+      "learning_rate": 1.1991999999999999e-08,
+      "loss": 0.4663,
+      "step": 247100
+    },
+    {
+      "epoch": 2.4715,
+      "grad_norm": 4.15252161026001,
+      "learning_rate": 1.1792e-08,
+      "loss": 0.4194,
+      "step": 247150
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 51.74559783935547,
+      "learning_rate": 1.1591999999999999e-08,
+      "loss": 0.3372,
+      "step": 247200
+    },
+    {
+      "epoch": 2.4725,
+      "grad_norm": 66.37777709960938,
+      "learning_rate": 1.1391999999999999e-08,
+      "loss": 0.3389,
+      "step": 247250
+    },
+    {
+      "epoch": 2.473,
+      "grad_norm": 105.33214569091797,
+      "learning_rate": 1.1195999999999998e-08,
+      "loss": 0.4773,
+      "step": 247300
+    },
+    {
+      "epoch": 2.4735,
+      "grad_norm": 44.509883880615234,
+      "learning_rate": 1.0996e-08,
+      "loss": 0.3653,
+      "step": 247350
+    },
+    {
+      "epoch": 2.474,
+      "grad_norm": 85.23392486572266,
+      "learning_rate": 1.0796e-08,
+      "loss": 0.3496,
+      "step": 247400
+    },
+    {
+      "epoch": 2.4745,
+      "grad_norm": 91.08809661865234,
+      "learning_rate": 1.0596e-08,
+      "loss": 0.451,
+      "step": 247450
+    },
+    {
+      "epoch": 2.475,
+      "grad_norm": 51.84107208251953,
+      "learning_rate": 1.0396000000000001e-08,
+      "loss": 0.4562,
+      "step": 247500
+    },
+    {
+      "epoch": 2.4755,
+      "grad_norm": 89.71752166748047,
+      "learning_rate": 1.0195999999999999e-08,
+      "loss": 0.3289,
+      "step": 247550
+    },
+    {
+      "epoch": 2.476,
+      "grad_norm": 30.12227439880371,
+      "learning_rate": 9.995999999999999e-09,
+      "loss": 0.3847,
+      "step": 247600
+    },
+    {
+      "epoch": 2.4765,
+      "grad_norm": 138.02243041992188,
+      "learning_rate": 9.795999999999999e-09,
+      "loss": 0.7013,
+      "step": 247650
+    },
+    {
+      "epoch": 2.477,
+      "grad_norm": 9.041874885559082,
+      "learning_rate": 9.596e-09,
+      "loss": 0.5115,
+      "step": 247700
+    },
+    {
+      "epoch": 2.4775,
+      "grad_norm": 2.327457904815674,
+      "learning_rate": 9.396e-09,
+      "loss": 0.3211,
+      "step": 247750
+    },
+    {
+      "epoch": 2.4779999999999998,
+      "grad_norm": 6.0508317947387695,
+      "learning_rate": 9.195999999999998e-09,
+      "loss": 0.5396,
+      "step": 247800
+    },
+    {
+      "epoch": 2.4785,
+      "grad_norm": 26.49738883972168,
+      "learning_rate": 8.996e-09,
+      "loss": 0.2972,
+      "step": 247850
+    },
+    {
+      "epoch": 2.479,
+      "grad_norm": 70.40245819091797,
+      "learning_rate": 8.796e-09,
+      "loss": 0.526,
+      "step": 247900
+    },
+    {
+      "epoch": 2.4795,
+      "grad_norm": 70.888916015625,
+      "learning_rate": 8.596e-09,
+      "loss": 0.4678,
+      "step": 247950
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.654665231704712,
+      "learning_rate": 8.396e-09,
+      "loss": 0.5367,
+      "step": 248000
+    },
+    {
+      "epoch": 2.4805,
+      "grad_norm": 16.878707885742188,
+      "learning_rate": 8.196e-09,
+      "loss": 0.2801,
+      "step": 248050
+    },
+    {
+      "epoch": 2.481,
+      "grad_norm": 47.633583068847656,
+      "learning_rate": 7.996e-09,
+      "loss": 0.3872,
+      "step": 248100
+    },
+    {
+      "epoch": 2.4815,
+      "grad_norm": 77.5364990234375,
+      "learning_rate": 7.795999999999999e-09,
+      "loss": 0.4372,
+      "step": 248150
+    },
+    {
+      "epoch": 2.482,
+      "grad_norm": 57.8331184387207,
+      "learning_rate": 7.596e-09,
+      "loss": 0.5912,
+      "step": 248200
+    },
+    {
+      "epoch": 2.4825,
+      "grad_norm": 64.51912689208984,
+      "learning_rate": 7.396e-09,
+      "loss": 0.4503,
+      "step": 248250
+    },
+    {
+      "epoch": 2.483,
+      "grad_norm": 10.555734634399414,
+      "learning_rate": 7.1959999999999996e-09,
+      "loss": 0.4095,
+      "step": 248300
+    },
+    {
+      "epoch": 2.4835,
+      "grad_norm": 58.91648483276367,
+      "learning_rate": 6.9959999999999994e-09,
+      "loss": 0.4893,
+      "step": 248350
+    },
+    {
+      "epoch": 2.484,
+      "grad_norm": 81.10791778564453,
+      "learning_rate": 6.795999999999999e-09,
+      "loss": 0.3603,
+      "step": 248400
+    },
+    {
+      "epoch": 2.4845,
+      "grad_norm": 69.9582748413086,
+      "learning_rate": 6.596e-09,
+      "loss": 0.3966,
+      "step": 248450
+    },
+    {
+      "epoch": 2.485,
+      "grad_norm": 2.605888843536377,
+      "learning_rate": 6.396e-09,
+      "loss": 0.4878,
+      "step": 248500
+    },
+    {
+      "epoch": 2.4855,
+      "grad_norm": 92.16374206542969,
+      "learning_rate": 6.196e-09,
+      "loss": 0.4327,
+      "step": 248550
+    },
+    {
+      "epoch": 2.4859999999999998,
+      "grad_norm": 90.9197998046875,
+      "learning_rate": 5.9959999999999996e-09,
+      "loss": 0.3959,
+      "step": 248600
+    },
+    {
+      "epoch": 2.4865,
+      "grad_norm": 76.15708923339844,
+      "learning_rate": 5.7959999999999994e-09,
+      "loss": 0.3849,
+      "step": 248650
+    },
+    {
+      "epoch": 2.487,
+      "grad_norm": 32.75033187866211,
+      "learning_rate": 5.596e-09,
+      "loss": 0.4641,
+      "step": 248700
+    },
+    {
+      "epoch": 2.4875,
+      "grad_norm": 103.26265716552734,
+      "learning_rate": 5.395999999999999e-09,
+      "loss": 0.3382,
+      "step": 248750
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 68.11605072021484,
+      "learning_rate": 5.196e-09,
+      "loss": 0.3231,
+      "step": 248800
+    },
+    {
+      "epoch": 2.4885,
+      "grad_norm": 4.257381439208984,
+      "learning_rate": 4.996e-09,
+      "loss": 0.4907,
+      "step": 248850
+    },
+    {
+      "epoch": 2.489,
+      "grad_norm": 33.47376251220703,
+      "learning_rate": 4.7959999999999996e-09,
+      "loss": 0.3289,
+      "step": 248900
+    },
+    {
+      "epoch": 2.4895,
+      "grad_norm": 49.842369079589844,
+      "learning_rate": 4.596e-09,
+      "loss": 0.234,
+      "step": 248950
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.662029504776001,
+      "learning_rate": 4.395999999999999e-09,
+      "loss": 0.3642,
+      "step": 249000
+    },
+    {
+      "epoch": 2.4905,
+      "grad_norm": 112.91320037841797,
+      "learning_rate": 4.196e-09,
+      "loss": 0.437,
+      "step": 249050
+    },
+    {
+      "epoch": 2.491,
+      "grad_norm": 2.2020812034606934,
+      "learning_rate": 3.996e-09,
+      "loss": 0.4311,
+      "step": 249100
+    },
+    {
+      "epoch": 2.4915,
+      "grad_norm": 20.80237579345703,
+      "learning_rate": 3.796e-09,
+      "loss": 0.4193,
+      "step": 249150
+    },
+    {
+      "epoch": 2.492,
+      "grad_norm": 47.812049865722656,
+      "learning_rate": 3.5959999999999996e-09,
+      "loss": 0.2837,
+      "step": 249200
+    },
+    {
+      "epoch": 2.4925,
+      "grad_norm": 14.877352714538574,
+      "learning_rate": 3.396e-09,
+      "loss": 0.4079,
+      "step": 249250
+    },
+    {
+      "epoch": 2.493,
+      "grad_norm": 31.925203323364258,
+      "learning_rate": 3.1959999999999997e-09,
+      "loss": 0.4154,
+      "step": 249300
+    },
+    {
+      "epoch": 2.4935,
+      "grad_norm": 7.781558513641357,
+      "learning_rate": 2.996e-09,
+      "loss": 0.3361,
+      "step": 249350
+    },
+    {
+      "epoch": 2.4939999999999998,
+      "grad_norm": 0.3294135630130768,
+      "learning_rate": 2.796e-09,
+      "loss": 0.4179,
+      "step": 249400
+    },
+    {
+      "epoch": 2.4945,
+      "grad_norm": 64.53218078613281,
+      "learning_rate": 2.5959999999999997e-09,
+      "loss": 0.4882,
+      "step": 249450
+    },
+    {
+      "epoch": 2.495,
+      "grad_norm": 44.542545318603516,
+      "learning_rate": 2.396e-09,
+      "loss": 0.3351,
+      "step": 249500
+    },
+    {
+      "epoch": 2.4955,
+      "grad_norm": 18.63250160217285,
+      "learning_rate": 2.196e-09,
+      "loss": 0.4181,
+      "step": 249550
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 42.67049789428711,
+      "learning_rate": 1.9959999999999997e-09,
+      "loss": 0.4524,
+      "step": 249600
+    },
+    {
+      "epoch": 2.4965,
+      "grad_norm": 4.257511615753174,
+      "learning_rate": 1.796e-09,
+      "loss": 0.3714,
+      "step": 249650
+    },
+    {
+      "epoch": 2.497,
+      "grad_norm": 38.738487243652344,
+      "learning_rate": 1.5959999999999999e-09,
+      "loss": 0.3591,
+      "step": 249700
+    },
+    {
+      "epoch": 2.4975,
+      "grad_norm": 50.39963150024414,
+      "learning_rate": 1.3960000000000001e-09,
+      "loss": 0.3474,
+      "step": 249750
+    },
+    {
+      "epoch": 2.498,
+      "grad_norm": 69.15322875976562,
+      "learning_rate": 1.196e-09,
+      "loss": 0.5561,
+      "step": 249800
+    },
+    {
+      "epoch": 2.4985,
+      "grad_norm": 7.292196273803711,
+      "learning_rate": 9.959999999999999e-10,
+      "loss": 0.5342,
+      "step": 249850
+    },
+    {
+      "epoch": 2.499,
+      "grad_norm": 1.3633363246917725,
+      "learning_rate": 7.96e-10,
+      "loss": 0.4622,
+      "step": 249900
+    },
+    {
+      "epoch": 2.4995,
+      "grad_norm": 82.1470947265625,
+      "learning_rate": 5.959999999999999e-10,
+      "loss": 0.5702,
+      "step": 249950
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 84.26116943359375,
+      "learning_rate": 3.96e-10,
+      "loss": 0.4723,
+      "step": 250000
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 250000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}