diff --git "a/checkpoint-13000/trainer_state.json" "b/checkpoint-13000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-13000/trainer_state.json"
@@ -0,0 +1,9285 @@
+{
+  "best_metric": 0.3308734893798828,
+  "best_model_checkpoint": "results_mt5XLSum_augmented/checkpoint-13000",
+  "epoch": 9.285714285714286,
+  "eval_steps": 1000,
+  "global_step": 13000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.007142857142857143,
+      "grad_norm": 10.661067962646484,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.8382,
+      "step": 10
+    },
+    {
+      "epoch": 0.014285714285714285,
+      "grad_norm": 11.73471450805664,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.6838,
+      "step": 20
+    },
+    {
+      "epoch": 0.02142857142857143,
+      "grad_norm": 13.098968505859375,
+      "learning_rate": 3e-06,
+      "loss": 1.86,
+      "step": 30
+    },
+    {
+      "epoch": 0.02857142857142857,
+      "grad_norm": 10.79481315612793,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.7851,
+      "step": 40
+    },
+    {
+      "epoch": 0.03571428571428571,
+      "grad_norm": 11.62800121307373,
+      "learning_rate": 5e-06,
+      "loss": 1.4648,
+      "step": 50
+    },
+    {
+      "epoch": 0.04285714285714286,
+      "grad_norm": 9.00180721282959,
+      "learning_rate": 6e-06,
+      "loss": 1.4355,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.218201637268066,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 1.3977,
+      "step": 70
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 7.159872531890869,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.9774,
+      "step": 80
+    },
+    {
+      "epoch": 0.06428571428571428,
+      "grad_norm": 6.163649559020996,
+      "learning_rate": 9e-06,
+      "loss": 0.8556,
+      "step": 90
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 4.800461292266846,
+      "learning_rate": 1e-05,
+      "loss": 0.8627,
+      "step": 100
+    },
+    {
+      "epoch": 0.07857142857142857,
+      "grad_norm": 4.373474597930908,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.7674,
+      "step": 110
+    },
+    {
+      "epoch": 0.08571428571428572,
+      "grad_norm": 4.36292839050293,
+      "learning_rate": 1.2e-05,
+      "loss": 0.7035,
+      "step": 120
+    },
+    {
+      "epoch": 0.09285714285714286,
+      "grad_norm": 4.634104251861572,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.9197,
+      "step": 130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.442883491516113,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.7712,
+      "step": 140
+    },
+    {
+      "epoch": 0.10714285714285714,
+      "grad_norm": 3.7063419818878174,
+      "learning_rate": 1.5e-05,
+      "loss": 0.8602,
+      "step": 150
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 3.7267696857452393,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.6758,
+      "step": 160
+    },
+    {
+      "epoch": 0.12142857142857143,
+      "grad_norm": 3.7582225799560547,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.8091,
+      "step": 170
+    },
+    {
+      "epoch": 0.12857142857142856,
+      "grad_norm": 2.829885482788086,
+      "learning_rate": 1.8e-05,
+      "loss": 0.8014,
+      "step": 180
+    },
+    {
+      "epoch": 0.1357142857142857,
+      "grad_norm": 3.4555258750915527,
+      "learning_rate": 1.9e-05,
+      "loss": 0.5562,
+      "step": 190
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 3.120464563369751,
+      "learning_rate": 2e-05,
+      "loss": 0.6391,
+      "step": 200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8185417652130127,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6501,
+      "step": 210
+    },
+    {
+      "epoch": 0.15714285714285714,
+      "grad_norm": 2.7110323905944824,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.9029,
+      "step": 220
+    },
+    {
+      "epoch": 0.16428571428571428,
+      "grad_norm": 3.336864709854126,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.6938,
+      "step": 230
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 2.9769392013549805,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6322,
+      "step": 240
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 2.5426135063171387,
+      "learning_rate": 2.5e-05,
+      "loss": 0.752,
+      "step": 250
+    },
+    {
+      "epoch": 0.18571428571428572,
+      "grad_norm": 3.2473714351654053,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.5993,
+      "step": 260
+    },
+    {
+      "epoch": 0.19285714285714287,
+      "grad_norm": 2.9979186058044434,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.5928,
+      "step": 270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1635003089904785,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.4335,
+      "step": 280
+    },
+    {
+      "epoch": 0.20714285714285716,
+      "grad_norm": 4.114761829376221,
+      "learning_rate": 2.9e-05,
+      "loss": 0.5023,
+      "step": 290
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 2.0567097663879395,
+      "learning_rate": 3e-05,
+      "loss": 0.5124,
+      "step": 300
+    },
+    {
+      "epoch": 0.22142857142857142,
+      "grad_norm": 3.0209622383117676,
+      "learning_rate": 3.1e-05,
+      "loss": 0.5092,
+      "step": 310
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 1.8497462272644043,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.6075,
+      "step": 320
+    },
+    {
+      "epoch": 0.2357142857142857,
+      "grad_norm": 1.6237268447875977,
+      "learning_rate": 3.3e-05,
+      "loss": 0.5343,
+      "step": 330
+    },
+    {
+      "epoch": 0.24285714285714285,
+      "grad_norm": 2.9820289611816406,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.5516,
+      "step": 340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.676515817642212,
+      "learning_rate": 3.5e-05,
+      "loss": 0.6621,
+      "step": 350
+    },
+    {
+      "epoch": 0.2571428571428571,
+      "grad_norm": 3.4376354217529297,
+      "learning_rate": 3.6e-05,
+      "loss": 0.4768,
+      "step": 360
+    },
+    {
+      "epoch": 0.2642857142857143,
+      "grad_norm": 3.2355964183807373,
+      "learning_rate": 3.7e-05,
+      "loss": 0.6184,
+      "step": 370
+    },
+    {
+      "epoch": 0.2714285714285714,
+      "grad_norm": 2.2971713542938232,
+      "learning_rate": 3.8e-05,
+      "loss": 0.7827,
+      "step": 380
+    },
+    {
+      "epoch": 0.2785714285714286,
+      "grad_norm": 2.442052125930786,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.5901,
+      "step": 390
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 2.3172521591186523,
+      "learning_rate": 4e-05,
+      "loss": 0.6067,
+      "step": 400
+    },
+    {
+      "epoch": 0.29285714285714287,
+      "grad_norm": 2.06640887260437,
+      "learning_rate": 4.1e-05,
+      "loss": 0.6589,
+      "step": 410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.416149854660034,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6489,
+      "step": 420
+    },
+    {
+      "epoch": 0.30714285714285716,
+      "grad_norm": 2.340235471725464,
+      "learning_rate": 4.3e-05,
+      "loss": 0.8339,
+      "step": 430
+    },
+    {
+      "epoch": 0.3142857142857143,
+      "grad_norm": 1.9825040102005005,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.7415,
+      "step": 440
+    },
+    {
+      "epoch": 0.32142857142857145,
+      "grad_norm": 1.8823323249816895,
+      "learning_rate": 4.5e-05,
+      "loss": 0.611,
+      "step": 450
+    },
+    {
+      "epoch": 0.32857142857142857,
+      "grad_norm": 2.3207123279571533,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.6172,
+      "step": 460
+    },
+    {
+      "epoch": 0.3357142857142857,
+      "grad_norm": 1.2963736057281494,
+      "learning_rate": 4.7e-05,
+      "loss": 0.4114,
+      "step": 470
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 2.191009044647217,
+      "learning_rate": 4.8e-05,
+      "loss": 0.6251,
+      "step": 480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1893374919891357,
+      "learning_rate": 4.9e-05,
+      "loss": 0.5966,
+      "step": 490
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 2.214414596557617,
+      "learning_rate": 5e-05,
+      "loss": 0.7495,
+      "step": 500
+    },
+    {
+      "epoch": 0.36428571428571427,
+      "grad_norm": 1.8343987464904785,
+      "learning_rate": 4.9962962962962964e-05,
+      "loss": 0.6179,
+      "step": 510
+    },
+    {
+      "epoch": 0.37142857142857144,
+      "grad_norm": 2.675177574157715,
+      "learning_rate": 4.9925925925925926e-05,
+      "loss": 0.7262,
+      "step": 520
+    },
+    {
+      "epoch": 0.37857142857142856,
+      "grad_norm": 1.7133303880691528,
+      "learning_rate": 4.9888888888888894e-05,
+      "loss": 0.546,
+      "step": 530
+    },
+    {
+      "epoch": 0.38571428571428573,
+      "grad_norm": 1.4926049709320068,
+      "learning_rate": 4.9851851851851855e-05,
+      "loss": 0.4057,
+      "step": 540
+    },
+    {
+      "epoch": 0.39285714285714285,
+      "grad_norm": 2.0434134006500244,
+      "learning_rate": 4.981481481481482e-05,
+      "loss": 1.014,
+      "step": 550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.037074089050293,
+      "learning_rate": 4.977777777777778e-05,
+      "loss": 0.6088,
+      "step": 560
+    },
+    {
+      "epoch": 0.40714285714285714,
+      "grad_norm": 2.6729607582092285,
+      "learning_rate": 4.974074074074074e-05,
+      "loss": 0.5986,
+      "step": 570
+    },
+    {
+      "epoch": 0.4142857142857143,
+      "grad_norm": 1.8161852359771729,
+      "learning_rate": 4.970370370370371e-05,
+      "loss": 0.4681,
+      "step": 580
+    },
+    {
+      "epoch": 0.42142857142857143,
+      "grad_norm": 2.140554666519165,
+      "learning_rate": 4.966666666666667e-05,
+      "loss": 0.5865,
+      "step": 590
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 2.3027639389038086,
+      "learning_rate": 4.962962962962963e-05,
+      "loss": 0.4166,
+      "step": 600
+    },
+    {
+      "epoch": 0.4357142857142857,
+      "grad_norm": 2.973132371902466,
+      "learning_rate": 4.959259259259259e-05,
+      "loss": 0.6394,
+      "step": 610
+    },
+    {
+      "epoch": 0.44285714285714284,
+      "grad_norm": 2.898897886276245,
+      "learning_rate": 4.955555555555556e-05,
+      "loss": 0.5572,
+      "step": 620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.100752353668213,
+      "learning_rate": 4.951851851851852e-05,
+      "loss": 0.4788,
+      "step": 630
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 2.4735984802246094,
+      "learning_rate": 4.9481481481481485e-05,
+      "loss": 0.756,
+      "step": 640
+    },
+    {
+      "epoch": 0.4642857142857143,
+      "grad_norm": 1.5895243883132935,
+      "learning_rate": 4.9444444444444446e-05,
+      "loss": 0.5265,
+      "step": 650
+    },
+    {
+      "epoch": 0.4714285714285714,
+      "grad_norm": 2.067650079727173,
+      "learning_rate": 4.940740740740741e-05,
+      "loss": 0.6079,
+      "step": 660
+    },
+    {
+      "epoch": 0.4785714285714286,
+      "grad_norm": 1.6676874160766602,
+      "learning_rate": 4.937037037037037e-05,
+      "loss": 0.5196,
+      "step": 670
+    },
+    {
+      "epoch": 0.4857142857142857,
+      "grad_norm": 1.6084502935409546,
+      "learning_rate": 4.933333333333334e-05,
+      "loss": 0.431,
+      "step": 680
+    },
+    {
+      "epoch": 0.4928571428571429,
+      "grad_norm": 2.8858065605163574,
+      "learning_rate": 4.92962962962963e-05,
+      "loss": 0.5329,
+      "step": 690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.657158613204956,
+      "learning_rate": 4.925925925925926e-05,
+      "loss": 0.5092,
+      "step": 700
+    },
+    {
+      "epoch": 0.5071428571428571,
+      "grad_norm": 2.636237144470215,
+      "learning_rate": 4.922222222222222e-05,
+      "loss": 0.4074,
+      "step": 710
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 2.5960123538970947,
+      "learning_rate": 4.918518518518519e-05,
+      "loss": 0.424,
+      "step": 720
+    },
+    {
+      "epoch": 0.5214285714285715,
+      "grad_norm": 1.7363989353179932,
+      "learning_rate": 4.9148148148148145e-05,
+      "loss": 0.7226,
+      "step": 730
+    },
+    {
+      "epoch": 0.5285714285714286,
+      "grad_norm": 2.8367726802825928,
+      "learning_rate": 4.9111111111111114e-05,
+      "loss": 0.5535,
+      "step": 740
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 2.1372838020324707,
+      "learning_rate": 4.9074074074074075e-05,
+      "loss": 0.6546,
+      "step": 750
+    },
+    {
+      "epoch": 0.5428571428571428,
+      "grad_norm": 1.9456530809402466,
+      "learning_rate": 4.903703703703704e-05,
+      "loss": 0.5718,
+      "step": 760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.146430015563965,
+      "learning_rate": 4.9e-05,
+      "loss": 0.486,
+      "step": 770
+    },
+    {
+      "epoch": 0.5571428571428572,
+      "grad_norm": 1.633537769317627,
+      "learning_rate": 4.896296296296297e-05,
+      "loss": 0.4602,
+      "step": 780
+    },
+    {
+      "epoch": 0.5642857142857143,
+      "grad_norm": 3.580615282058716,
+      "learning_rate": 4.892592592592593e-05,
+      "loss": 0.5991,
+      "step": 790
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 2.354482889175415,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.639,
+      "step": 800
+    },
+    {
+      "epoch": 0.5785714285714286,
+      "grad_norm": 1.701165795326233,
+      "learning_rate": 4.885185185185185e-05,
+      "loss": 0.4895,
+      "step": 810
+    },
+    {
+      "epoch": 0.5857142857142857,
+      "grad_norm": 1.7530277967453003,
+      "learning_rate": 4.881481481481482e-05,
+      "loss": 0.5029,
+      "step": 820
+    },
+    {
+      "epoch": 0.5928571428571429,
+      "grad_norm": 1.4377954006195068,
+      "learning_rate": 4.8777777777777775e-05,
+      "loss": 0.4668,
+      "step": 830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9733954668045044,
+      "learning_rate": 4.874074074074074e-05,
+      "loss": 0.6434,
+      "step": 840
+    },
+    {
+      "epoch": 0.6071428571428571,
+      "grad_norm": 1.5659862756729126,
+      "learning_rate": 4.8703703703703704e-05,
+      "loss": 0.4719,
+      "step": 850
+    },
+    {
+      "epoch": 0.6142857142857143,
+      "grad_norm": 1.9549959897994995,
+      "learning_rate": 4.866666666666667e-05,
+      "loss": 0.5003,
+      "step": 860
+    },
+    {
+      "epoch": 0.6214285714285714,
+      "grad_norm": 2.0998220443725586,
+      "learning_rate": 4.862962962962963e-05,
+      "loss": 0.4666,
+      "step": 870
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 1.6551822423934937,
+      "learning_rate": 4.8592592592592596e-05,
+      "loss": 0.5508,
+      "step": 880
+    },
+    {
+      "epoch": 0.6357142857142857,
+      "grad_norm": 2.268826723098755,
+      "learning_rate": 4.855555555555556e-05,
+      "loss": 0.5333,
+      "step": 890
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 2.170297861099243,
+      "learning_rate": 4.851851851851852e-05,
+      "loss": 0.4724,
+      "step": 900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3737900257110596,
+      "learning_rate": 4.848148148148148e-05,
+      "loss": 0.5938,
+      "step": 910
+    },
+    {
+      "epoch": 0.6571428571428571,
+      "grad_norm": 1.5697389841079712,
+      "learning_rate": 4.844444444444445e-05,
+      "loss": 0.357,
+      "step": 920
+    },
+    {
+      "epoch": 0.6642857142857143,
+      "grad_norm": 1.4354273080825806,
+      "learning_rate": 4.840740740740741e-05,
+      "loss": 0.2648,
+      "step": 930
+    },
+    {
+      "epoch": 0.6714285714285714,
+      "grad_norm": 1.1631938219070435,
+      "learning_rate": 4.837037037037037e-05,
+      "loss": 0.4647,
+      "step": 940
+    },
+    {
+      "epoch": 0.6785714285714286,
+      "grad_norm": 2.594999313354492,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 0.6831,
+      "step": 950
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 2.2979557514190674,
+      "learning_rate": 4.82962962962963e-05,
+      "loss": 0.4363,
+      "step": 960
+    },
+    {
+      "epoch": 0.6928571428571428,
+      "grad_norm": 3.0777416229248047,
+      "learning_rate": 4.825925925925926e-05,
+      "loss": 0.5995,
+      "step": 970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.430807113647461,
+      "learning_rate": 4.8222222222222225e-05,
+      "loss": 0.6433,
+      "step": 980
+    },
+    {
+      "epoch": 0.7071428571428572,
+      "grad_norm": 1.7465846538543701,
+      "learning_rate": 4.818518518518519e-05,
+      "loss": 0.4973,
+      "step": 990
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 2.805053472518921,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.556,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "eval_loss": 0.4978465139865875,
+      "eval_rouge1": 0.8844,
+      "eval_rouge2": 0.8183,
+      "eval_rougeL": 0.8811,
+      "eval_runtime": 122.1438,
+      "eval_samples_per_second": 11.462,
+      "eval_steps_per_second": 5.731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7214285714285714,
+      "grad_norm": 1.6127879619598389,
+      "learning_rate": 4.811111111111111e-05,
+      "loss": 0.5748,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7285714285714285,
+      "grad_norm": 1.1071356534957886,
+      "learning_rate": 4.807407407407408e-05,
+      "loss": 0.5246,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7357142857142858,
+      "grad_norm": 1.9362713098526,
+      "learning_rate": 4.803703703703704e-05,
+      "loss": 0.563,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 2.037553548812866,
+      "learning_rate": 4.8e-05,
+      "loss": 0.466,
+      "step": 1040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.196617841720581,
+      "learning_rate": 4.796296296296296e-05,
+      "loss": 0.5647,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7571428571428571,
+      "grad_norm": 1.278428077697754,
+      "learning_rate": 4.792592592592593e-05,
+      "loss": 0.4821,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7642857142857142,
+      "grad_norm": 1.3506104946136475,
+      "learning_rate": 4.7888888888888886e-05,
+      "loss": 0.5194,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7714285714285715,
+      "grad_norm": 2.3870656490325928,
+      "learning_rate": 4.7851851851851854e-05,
+      "loss": 0.7373,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7785714285714286,
+      "grad_norm": 2.071242094039917,
+      "learning_rate": 4.7814814814814816e-05,
+      "loss": 0.5598,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 1.8460086584091187,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 0.6184,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7928571428571428,
+      "grad_norm": 3.804724931716919,
+      "learning_rate": 4.774074074074074e-05,
+      "loss": 0.5978,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.614772081375122,
+      "learning_rate": 4.770370370370371e-05,
+      "loss": 0.6203,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8071428571428572,
+      "grad_norm": 2.068122386932373,
+      "learning_rate": 4.766666666666667e-05,
+      "loss": 0.6149,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8142857142857143,
+      "grad_norm": 1.675881266593933,
+      "learning_rate": 4.762962962962963e-05,
+      "loss": 0.4437,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8214285714285714,
+      "grad_norm": 1.865435004234314,
+      "learning_rate": 4.759259259259259e-05,
+      "loss": 0.5166,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8285714285714286,
+      "grad_norm": 1.0480509996414185,
+      "learning_rate": 4.755555555555556e-05,
+      "loss": 0.3827,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8357142857142857,
+      "grad_norm": 2.218554735183716,
+      "learning_rate": 4.751851851851852e-05,
+      "loss": 0.6641,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8428571428571429,
+      "grad_norm": 2.510831832885742,
+      "learning_rate": 4.7481481481481483e-05,
+      "loss": 0.496,
+      "step": 1180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8328824043273926,
+      "learning_rate": 4.7444444444444445e-05,
+      "loss": 0.5466,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.5480728149414062,
+      "learning_rate": 4.740740740740741e-05,
+      "loss": 0.4992,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8642857142857143,
+      "grad_norm": 1.3723492622375488,
+      "learning_rate": 4.737037037037037e-05,
+      "loss": 0.5014,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8714285714285714,
+      "grad_norm": 1.7510666847229004,
+      "learning_rate": 4.7333333333333336e-05,
+      "loss": 0.5471,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8785714285714286,
+      "grad_norm": 2.2057995796203613,
+      "learning_rate": 4.72962962962963e-05,
+      "loss": 0.6142,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8857142857142857,
+      "grad_norm": 1.7922954559326172,
+      "learning_rate": 4.7259259259259266e-05,
+      "loss": 0.5199,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 1.9541053771972656,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 0.44,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2869590520858765,
+      "learning_rate": 4.718518518518519e-05,
+      "loss": 0.5157,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9071428571428571,
+      "grad_norm": 1.7564722299575806,
+      "learning_rate": 4.714814814814815e-05,
+      "loss": 0.4985,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 1.0782675743103027,
+      "learning_rate": 4.711111111111111e-05,
+      "loss": 0.3195,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9214285714285714,
+      "grad_norm": 1.7535449266433716,
+      "learning_rate": 4.7074074074074074e-05,
+      "loss": 0.376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 1.581485629081726,
+      "learning_rate": 4.703703703703704e-05,
+      "loss": 0.5975,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9357142857142857,
+      "grad_norm": 2.739900827407837,
+      "learning_rate": 4.7e-05,
+      "loss": 0.457,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9428571428571428,
+      "grad_norm": 2.382187604904175,
+      "learning_rate": 4.6962962962962966e-05,
+      "loss": 0.5424,
+      "step": 1320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.75946843624115,
+      "learning_rate": 4.692592592592593e-05,
+      "loss": 0.3563,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9571428571428572,
+      "grad_norm": 1.8159079551696777,
+      "learning_rate": 4.6888888888888895e-05,
+      "loss": 0.4502,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9642857142857143,
+      "grad_norm": 2.605283260345459,
+      "learning_rate": 4.685185185185185e-05,
+      "loss": 0.4779,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 2.594231605529785,
+      "learning_rate": 4.681481481481482e-05,
+      "loss": 0.4901,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9785714285714285,
+      "grad_norm": 2.109367609024048,
+      "learning_rate": 4.677777777777778e-05,
+      "loss": 0.5378,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9857142857142858,
+      "grad_norm": 1.960496425628662,
+      "learning_rate": 4.674074074074074e-05,
+      "loss": 0.6129,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9928571428571429,
+      "grad_norm": 3.4135870933532715,
+      "learning_rate": 4.67037037037037e-05,
+      "loss": 0.7069,
+      "step": 1390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.441308617591858,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.4686,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0071428571428571,
+      "grad_norm": 1.9842432737350464,
+      "learning_rate": 4.662962962962963e-05,
+      "loss": 0.604,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0142857142857142,
+      "grad_norm": 1.3867950439453125,
+      "learning_rate": 4.6592592592592595e-05,
+      "loss": 0.4168,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0214285714285714,
+      "grad_norm": 2.118037462234497,
+      "learning_rate": 4.6555555555555556e-05,
+      "loss": 0.6484,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0285714285714285,
+      "grad_norm": 1.4064522981643677,
+      "learning_rate": 4.6518518518518525e-05,
+      "loss": 0.5275,
+      "step": 1440
+    },
+    {
+      "epoch": 1.0357142857142858,
+      "grad_norm": 2.644491672515869,
+      "learning_rate": 4.648148148148148e-05,
+      "loss": 0.5361,
+      "step": 1450
+    },
+    {
+      "epoch": 1.042857142857143,
+      "grad_norm": 1.4005937576293945,
+      "learning_rate": 4.644444444444445e-05,
+      "loss": 0.4497,
+      "step": 1460
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.773334264755249,
+      "learning_rate": 4.640740740740741e-05,
+      "loss": 0.4372,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0571428571428572,
+      "grad_norm": 2.1667587757110596,
+      "learning_rate": 4.637037037037038e-05,
+      "loss": 0.5211,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0642857142857143,
+      "grad_norm": 1.1993277072906494,
+      "learning_rate": 4.633333333333333e-05,
+      "loss": 0.3694,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 1.5526480674743652,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.686,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0785714285714285,
+      "grad_norm": 1.5041449069976807,
+      "learning_rate": 4.625925925925926e-05,
+      "loss": 0.4536,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0857142857142856,
+      "grad_norm": 1.719254970550537,
+      "learning_rate": 4.6222222222222224e-05,
+      "loss": 0.4687,
+      "step": 1520
+    },
+    {
+      "epoch": 1.092857142857143,
+      "grad_norm": 1.9565083980560303,
+      "learning_rate": 4.6185185185185185e-05,
+      "loss": 0.4054,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.2271467447280884,
+      "learning_rate": 4.6148148148148154e-05,
+      "loss": 0.4189,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1071428571428572,
+      "grad_norm": 1.731244683265686,
+      "learning_rate": 4.6111111111111115e-05,
+      "loss": 0.4519,
+      "step": 1550
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 1.3039075136184692,
+      "learning_rate": 4.607407407407408e-05,
+      "loss": 0.3911,
+      "step": 1560
+    },
+    {
+      "epoch": 1.1214285714285714,
+      "grad_norm": 1.3420417308807373,
+      "learning_rate": 4.603703703703704e-05,
+      "loss": 0.4239,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1285714285714286,
+      "grad_norm": 2.2307205200195312,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.4675,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1357142857142857,
+      "grad_norm": 2.384147882461548,
+      "learning_rate": 4.596296296296296e-05,
+      "loss": 0.3963,
+      "step": 1590
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 1.6016713380813599,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.4561,
+      "step": 1600
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.4093197584152222,
+      "learning_rate": 4.588888888888889e-05,
+      "loss": 0.4708,
+      "step": 1610
+    },
+    {
+      "epoch": 1.157142857142857,
+      "grad_norm": 1.9773272275924683,
+      "learning_rate": 4.585185185185185e-05,
+      "loss": 0.5259,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1642857142857144,
+      "grad_norm": 1.169757604598999,
+      "learning_rate": 4.5814814814814815e-05,
+      "loss": 0.3413,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1714285714285715,
+      "grad_norm": 2.1033947467803955,
+      "learning_rate": 4.577777777777778e-05,
+      "loss": 0.4888,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1785714285714286,
+      "grad_norm": 1.2455283403396606,
+      "learning_rate": 4.5740740740740745e-05,
+      "loss": 0.5935,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1857142857142857,
+      "grad_norm": 1.283308982849121,
+      "learning_rate": 4.5703703703703706e-05,
+      "loss": 0.3946,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1928571428571428,
+      "grad_norm": 1.9639955759048462,
+      "learning_rate": 4.566666666666667e-05,
+      "loss": 0.519,
+      "step": 1670
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9380689263343811,
+      "learning_rate": 4.5629629629629636e-05,
+      "loss": 0.3357,
+      "step": 1680
+    },
+    {
+      "epoch": 1.207142857142857,
+      "grad_norm": 2.330310344696045,
+      "learning_rate": 4.559259259259259e-05,
+      "loss": 0.5135,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 1.5911920070648193,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 0.4165,
+      "step": 1700
+    },
+    {
+      "epoch": 1.2214285714285715,
+      "grad_norm": 1.7522234916687012,
+      "learning_rate": 4.551851851851852e-05,
+      "loss": 0.5797,
+      "step": 1710
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 2.265571355819702,
+      "learning_rate": 4.548148148148149e-05,
+      "loss": 0.3943,
+      "step": 1720
+    },
+    {
+      "epoch": 1.2357142857142858,
+      "grad_norm": 2.530675172805786,
+      "learning_rate": 4.5444444444444444e-05,
+      "loss": 0.6279,
+      "step": 1730
+    },
+    {
+      "epoch": 1.2428571428571429,
+      "grad_norm": 2.072864055633545,
+      "learning_rate": 4.540740740740741e-05,
+      "loss": 0.512,
+      "step": 1740
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5505369901657104,
+      "learning_rate": 4.5370370370370374e-05,
+      "loss": 0.3494,
+      "step": 1750
+    },
+    {
+      "epoch": 1.2571428571428571,
+      "grad_norm": 1.9888116121292114,
+      "learning_rate": 4.5333333333333335e-05,
+      "loss": 0.5841,
+      "step": 1760
+    },
+    {
+      "epoch": 1.2642857142857142,
+      "grad_norm": 1.6056774854660034,
+      "learning_rate": 4.52962962962963e-05,
+      "loss": 0.5611,
+      "step": 1770
+    },
+    {
+      "epoch": 1.2714285714285714,
+      "grad_norm": 1.7950221300125122,
+      "learning_rate": 4.5259259259259265e-05,
+      "loss": 0.6097,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2785714285714285,
+      "grad_norm": 1.8906399011611938,
+      "learning_rate": 4.522222222222223e-05,
+      "loss": 0.4837,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 1.3988184928894043,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 0.4806,
+      "step": 1800
+    },
+    {
+      "epoch": 1.292857142857143,
+      "grad_norm": 1.160243272781372,
+      "learning_rate": 4.514814814814815e-05,
+      "loss": 0.5224,
+      "step": 1810
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.0152113437652588,
+      "learning_rate": 4.511111111111112e-05,
+      "loss": 0.4115,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3071428571428572,
+      "grad_norm": 1.6176999807357788,
+      "learning_rate": 4.507407407407407e-05,
+      "loss": 0.4458,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3142857142857143,
+      "grad_norm": 1.904784917831421,
+      "learning_rate": 4.503703703703704e-05,
+      "loss": 0.5552,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3214285714285714,
+      "grad_norm": 1.0539710521697998,
+      "learning_rate": 4.5e-05,
+      "loss": 0.438,
+      "step": 1850
+    },
+    {
+      "epoch": 1.3285714285714285,
+      "grad_norm": 1.3552178144454956,
+      "learning_rate": 4.496296296296297e-05,
+      "loss": 0.2862,
+      "step": 1860
+    },
+    {
+      "epoch": 1.3357142857142856,
+      "grad_norm": 1.3787767887115479,
+      "learning_rate": 4.4925925925925926e-05,
+      "loss": 0.5173,
+      "step": 1870
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 2.570422649383545,
+      "learning_rate": 4.4888888888888894e-05,
+      "loss": 0.4581,
+      "step": 1880
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5974104404449463,
+      "learning_rate": 4.4851851851851856e-05,
+      "loss": 0.4599,
+      "step": 1890
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 1.4105775356292725,
+      "learning_rate": 4.481481481481482e-05,
+      "loss": 0.3823,
+      "step": 1900
+    },
+    {
+      "epoch": 1.3642857142857143,
+      "grad_norm": 2.1751532554626465,
+      "learning_rate": 4.477777777777778e-05,
+      "loss": 0.4421,
+      "step": 1910
+    },
+    {
+      "epoch": 1.3714285714285714,
+      "grad_norm": 1.9956297874450684,
+      "learning_rate": 4.474074074074075e-05,
+      "loss": 0.4082,
+      "step": 1920
+    },
+    {
+      "epoch": 1.3785714285714286,
+      "grad_norm": 1.6159803867340088,
+      "learning_rate": 4.47037037037037e-05,
+      "loss": 0.3961,
+      "step": 1930
+    },
+    {
+      "epoch": 1.3857142857142857,
+      "grad_norm": 1.4909430742263794,
+      "learning_rate": 4.466666666666667e-05,
+      "loss": 0.4635,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3928571428571428,
+      "grad_norm": 1.5630055665969849,
+      "learning_rate": 4.462962962962963e-05,
+      "loss": 0.5968,
+      "step": 1950
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2496933937072754,
+      "learning_rate": 4.4592592592592594e-05,
+      "loss": 0.4546,
+      "step": 1960
+    },
+    {
+      "epoch": 1.407142857142857,
+      "grad_norm": 1.6497224569320679,
+      "learning_rate": 4.4555555555555555e-05,
+      "loss": 0.354,
+      "step": 1970
+    },
+    {
+      "epoch": 1.4142857142857144,
+      "grad_norm": 2.069955587387085,
+      "learning_rate": 4.4518518518518523e-05,
+      "loss": 0.4388,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4214285714285715,
+      "grad_norm": 1.6338075399398804,
+      "learning_rate": 4.4481481481481485e-05,
+      "loss": 0.5459,
+      "step": 1990
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 1.3558902740478516,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.3139,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "eval_loss": 0.43829917907714844,
+      "eval_rouge1": 0.8911,
+      "eval_rouge2": 0.8251,
+      "eval_rougeL": 0.8873,
+      "eval_runtime": 121.8873,
+      "eval_samples_per_second": 11.486,
+      "eval_steps_per_second": 5.743,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4357142857142857,
+      "grad_norm": 1.8123821020126343,
+      "learning_rate": 4.440740740740741e-05,
+      "loss": 0.5864,
+      "step": 2010
+    },
+    {
+      "epoch": 1.4428571428571428,
+      "grad_norm": 0.8494770526885986,
+      "learning_rate": 4.4370370370370376e-05,
+      "loss": 0.3284,
+      "step": 2020
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2536141872406006,
+      "learning_rate": 4.433333333333334e-05,
+      "loss": 0.3738,
+      "step": 2030
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 2.971925735473633,
+      "learning_rate": 4.42962962962963e-05,
+      "loss": 0.5294,
+      "step": 2040
+    },
+    {
+      "epoch": 1.4642857142857144,
+      "grad_norm": 1.7820425033569336,
+      "learning_rate": 4.425925925925926e-05,
+      "loss": 0.505,
+      "step": 2050
+    },
+    {
+      "epoch": 1.4714285714285715,
+      "grad_norm": 1.196044683456421,
+      "learning_rate": 4.422222222222222e-05,
+      "loss": 0.3921,
+      "step": 2060
+    },
+    {
+      "epoch": 1.4785714285714286,
+      "grad_norm": 0.9053621888160706,
+      "learning_rate": 4.4185185185185184e-05,
+      "loss": 0.2386,
+      "step": 2070
+    },
+    {
+      "epoch": 1.4857142857142858,
+      "grad_norm": 1.8388108015060425,
+      "learning_rate": 4.414814814814815e-05,
+      "loss": 0.4309,
+      "step": 2080
+    },
+    {
+      "epoch": 1.4928571428571429,
+      "grad_norm": 2.25136137008667,
+      "learning_rate": 4.4111111111111114e-05,
+      "loss": 0.3918,
+      "step": 2090
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.960864782333374,
+      "learning_rate": 4.4074074074074076e-05,
+      "loss": 0.4754,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5071428571428571,
+      "grad_norm": 2.4653213024139404,
+      "learning_rate": 4.403703703703704e-05,
+      "loss": 0.4545,
+      "step": 2110
+    },
+    {
+      "epoch": 1.5142857142857142,
+      "grad_norm": 1.8694462776184082,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.3707,
+      "step": 2120
+    },
+    {
+      "epoch": 1.5214285714285714,
+      "grad_norm": 2.240447521209717,
+      "learning_rate": 4.396296296296297e-05,
+      "loss": 0.5174,
+      "step": 2130
+    },
+    {
+      "epoch": 1.5285714285714285,
+      "grad_norm": 0.8589600920677185,
+      "learning_rate": 4.392592592592593e-05,
+      "loss": 0.3802,
+      "step": 2140
+    },
+    {
+      "epoch": 1.5357142857142856,
+      "grad_norm": 2.495075225830078,
+      "learning_rate": 4.388888888888889e-05,
+      "loss": 0.4735,
+      "step": 2150
+    },
+    {
+      "epoch": 1.5428571428571427,
+      "grad_norm": 1.5384397506713867,
+      "learning_rate": 4.385185185185185e-05,
+      "loss": 0.4844,
+      "step": 2160
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2120758295059204,
+      "learning_rate": 4.381481481481482e-05,
+      "loss": 0.3029,
+      "step": 2170
+    },
+    {
+      "epoch": 1.5571428571428572,
+      "grad_norm": 2.0210671424865723,
+      "learning_rate": 4.377777777777778e-05,
+      "loss": 0.688,
+      "step": 2180
+    },
+    {
+      "epoch": 1.5642857142857143,
+      "grad_norm": 2.322322368621826,
+      "learning_rate": 4.374074074074074e-05,
+      "loss": 0.4673,
+      "step": 2190
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 1.8948960304260254,
+      "learning_rate": 4.3703703703703705e-05,
+      "loss": 0.3698,
+      "step": 2200
+    },
+    {
+      "epoch": 1.5785714285714287,
+      "grad_norm": 1.776141881942749,
+      "learning_rate": 4.3666666666666666e-05,
+      "loss": 0.3611,
+      "step": 2210
+    },
+    {
+      "epoch": 1.5857142857142859,
+      "grad_norm": 2.8628015518188477,
+      "learning_rate": 4.3629629629629635e-05,
+      "loss": 0.4504,
+      "step": 2220
+    },
+    {
+      "epoch": 1.592857142857143,
+      "grad_norm": 1.8579275608062744,
+      "learning_rate": 4.3592592592592596e-05,
+      "loss": 0.5131,
+      "step": 2230
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1070181131362915,
+      "learning_rate": 4.355555555555556e-05,
+      "loss": 0.4187,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6071428571428572,
+      "grad_norm": 1.3833059072494507,
+      "learning_rate": 4.351851851851852e-05,
+      "loss": 0.4301,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6142857142857143,
+      "grad_norm": 1.6870567798614502,
+      "learning_rate": 4.348148148148148e-05,
+      "loss": 0.5542,
+      "step": 2260
+    },
+    {
+      "epoch": 1.6214285714285714,
+      "grad_norm": 1.582582712173462,
+      "learning_rate": 4.344444444444445e-05,
+      "loss": 0.5192,
+      "step": 2270
+    },
+    {
+      "epoch": 1.6285714285714286,
+      "grad_norm": 3.3700509071350098,
+      "learning_rate": 4.340740740740741e-05,
+      "loss": 0.3268,
+      "step": 2280
+    },
+    {
+      "epoch": 1.6357142857142857,
+      "grad_norm": 3.0057899951934814,
+      "learning_rate": 4.337037037037037e-05,
+      "loss": 0.3787,
+      "step": 2290
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 1.302416205406189,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.4793,
+      "step": 2300
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.42720103263855,
+      "learning_rate": 4.3296296296296296e-05,
+      "loss": 0.4124,
+      "step": 2310
+    },
+    {
+      "epoch": 1.657142857142857,
+      "grad_norm": 1.455609917640686,
+      "learning_rate": 4.325925925925926e-05,
+      "loss": 0.272,
+      "step": 2320
+    },
+    {
+      "epoch": 1.6642857142857141,
+      "grad_norm": 2.1332924365997314,
+      "learning_rate": 4.3222222222222226e-05,
+      "loss": 0.4859,
+      "step": 2330
+    },
+    {
+      "epoch": 1.6714285714285713,
+      "grad_norm": 1.977156162261963,
+      "learning_rate": 4.318518518518519e-05,
+      "loss": 0.4017,
+      "step": 2340
+    },
+    {
+      "epoch": 1.6785714285714286,
+      "grad_norm": 1.7197158336639404,
+      "learning_rate": 4.314814814814815e-05,
+      "loss": 0.376,
+      "step": 2350
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.8615891933441162,
+      "learning_rate": 4.311111111111111e-05,
+      "loss": 0.3383,
+      "step": 2360
+    },
+    {
+      "epoch": 1.6928571428571428,
+      "grad_norm": 1.2501980066299438,
+      "learning_rate": 4.307407407407408e-05,
+      "loss": 0.3999,
+      "step": 2370
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7977019548416138,
+      "learning_rate": 4.303703703703704e-05,
+      "loss": 0.3424,
+      "step": 2380
+    },
+    {
+      "epoch": 1.7071428571428573,
+      "grad_norm": 2.265807867050171,
+      "learning_rate": 4.3e-05,
+      "loss": 0.4345,
+      "step": 2390
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 2.348353624343872,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.4395,
+      "step": 2400
+    },
+    {
+      "epoch": 1.7214285714285715,
+      "grad_norm": 2.585843801498413,
+      "learning_rate": 4.292592592592593e-05,
+      "loss": 0.5208,
+      "step": 2410
+    },
+    {
+      "epoch": 1.7285714285714286,
+      "grad_norm": 1.1487417221069336,
+      "learning_rate": 4.2888888888888886e-05,
+      "loss": 0.4635,
+      "step": 2420
+    },
+    {
+      "epoch": 1.7357142857142858,
+      "grad_norm": 1.206634521484375,
+      "learning_rate": 4.2851851851851855e-05,
+      "loss": 0.3521,
+      "step": 2430
+    },
+    {
+      "epoch": 1.7428571428571429,
+      "grad_norm": 2.136702299118042,
+      "learning_rate": 4.2814814814814816e-05,
+      "loss": 0.4582,
+      "step": 2440
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.2831017971038818,
+      "learning_rate": 4.277777777777778e-05,
+      "loss": 0.397,
+      "step": 2450
+    },
+    {
+      "epoch": 1.7571428571428571,
+      "grad_norm": 2.313405990600586,
+      "learning_rate": 4.274074074074074e-05,
+      "loss": 0.431,
+      "step": 2460
+    },
+    {
+      "epoch": 1.7642857142857142,
+      "grad_norm": 1.8922353982925415,
+      "learning_rate": 4.270370370370371e-05,
+      "loss": 0.4396,
+      "step": 2470
+    },
+    {
+      "epoch": 1.7714285714285714,
+      "grad_norm": 1.735303521156311,
+      "learning_rate": 4.266666666666667e-05,
+      "loss": 0.4019,
+      "step": 2480
+    },
+    {
+      "epoch": 1.7785714285714285,
+      "grad_norm": 1.1989376544952393,
+      "learning_rate": 4.262962962962963e-05,
+      "loss": 0.3317,
+      "step": 2490
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 1.709370732307434,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.4721,
+      "step": 2500
+    },
+    {
+      "epoch": 1.7928571428571427,
+      "grad_norm": 1.3655775785446167,
+      "learning_rate": 4.255555555555556e-05,
+      "loss": 0.464,
+      "step": 2510
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.2292691469192505,
+      "learning_rate": 4.2518518518518515e-05,
+      "loss": 0.3765,
+      "step": 2520
+    },
+    {
+      "epoch": 1.8071428571428572,
+      "grad_norm": 2.6490797996520996,
+      "learning_rate": 4.2481481481481484e-05,
+      "loss": 0.4989,
+      "step": 2530
+    },
+    {
+      "epoch": 1.8142857142857143,
+      "grad_norm": 1.8564647436141968,
+      "learning_rate": 4.2444444444444445e-05,
+      "loss": 0.3776,
+      "step": 2540
+    },
+    {
+      "epoch": 1.8214285714285714,
+      "grad_norm": 1.9681627750396729,
+      "learning_rate": 4.240740740740741e-05,
+      "loss": 0.4717,
+      "step": 2550
+    },
+    {
+      "epoch": 1.8285714285714287,
+      "grad_norm": 2.1326770782470703,
+      "learning_rate": 4.237037037037037e-05,
+      "loss": 0.3212,
+      "step": 2560
+    },
+    {
+      "epoch": 1.8357142857142859,
+      "grad_norm": 1.8122767210006714,
+      "learning_rate": 4.233333333333334e-05,
+      "loss": 0.3619,
+      "step": 2570
+    },
+    {
+      "epoch": 1.842857142857143,
+      "grad_norm": 1.4822399616241455,
+      "learning_rate": 4.22962962962963e-05,
+      "loss": 0.4631,
+      "step": 2580
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.278700828552246,
+      "learning_rate": 4.225925925925926e-05,
+      "loss": 0.3018,
+      "step": 2590
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 2.3148486614227295,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.4084,
+      "step": 2600
+    },
+    {
+      "epoch": 1.8642857142857143,
+      "grad_norm": 1.5277279615402222,
+      "learning_rate": 4.218518518518519e-05,
+      "loss": 0.5295,
+      "step": 2610
+    },
+    {
+      "epoch": 1.8714285714285714,
+      "grad_norm": 1.3603259325027466,
+      "learning_rate": 4.2148148148148145e-05,
+      "loss": 0.4727,
+      "step": 2620
+    },
+    {
+      "epoch": 1.8785714285714286,
+      "grad_norm": 1.9577744007110596,
+      "learning_rate": 4.211111111111111e-05,
+      "loss": 0.5915,
+      "step": 2630
+    },
+    {
+      "epoch": 1.8857142857142857,
+      "grad_norm": 1.0424437522888184,
+      "learning_rate": 4.2074074074074075e-05,
+      "loss": 0.3386,
+      "step": 2640
+    },
+    {
+      "epoch": 1.8928571428571428,
+      "grad_norm": 2.7555553913116455,
+      "learning_rate": 4.203703703703704e-05,
+      "loss": 0.5003,
+      "step": 2650
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.9913907051086426,
+      "learning_rate": 4.2e-05,
+      "loss": 0.3875,
+      "step": 2660
+    },
+    {
+      "epoch": 1.907142857142857,
+      "grad_norm": 1.8053233623504639,
+      "learning_rate": 4.1962962962962966e-05,
+      "loss": 0.2318,
+      "step": 2670
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 1.7686558961868286,
+      "learning_rate": 4.192592592592593e-05,
+      "loss": 0.3772,
+      "step": 2680
+    },
+    {
+      "epoch": 1.9214285714285713,
+      "grad_norm": 1.4202839136123657,
+      "learning_rate": 4.188888888888889e-05,
+      "loss": 0.3742,
+      "step": 2690
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 2.7964282035827637,
+      "learning_rate": 4.185185185185185e-05,
+      "loss": 0.2901,
+      "step": 2700
+    },
+    {
+      "epoch": 1.9357142857142857,
+      "grad_norm": 2.0525360107421875,
+      "learning_rate": 4.181481481481482e-05,
+      "loss": 0.3146,
+      "step": 2710
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "grad_norm": 1.937103033065796,
+      "learning_rate": 4.177777777777778e-05,
+      "loss": 0.3871,
+      "step": 2720
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.1534152030944824,
+      "learning_rate": 4.174074074074074e-05,
+      "loss": 0.5309,
+      "step": 2730
+    },
+    {
+      "epoch": 1.9571428571428573,
+      "grad_norm": 1.60648512840271,
+      "learning_rate": 4.1703703703703704e-05,
+      "loss": 0.3859,
+      "step": 2740
+    },
+    {
+      "epoch": 1.9642857142857144,
+      "grad_norm": 2.2782654762268066,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.3859,
+      "step": 2750
+    },
+    {
+      "epoch": 1.9714285714285715,
+      "grad_norm": 1.9134782552719116,
+      "learning_rate": 4.162962962962963e-05,
+      "loss": 0.565,
+      "step": 2760
+    },
+    {
+      "epoch": 1.9785714285714286,
+      "grad_norm": 1.4029120206832886,
+      "learning_rate": 4.1592592592592595e-05,
+      "loss": 0.3373,
+      "step": 2770
+    },
+    {
+      "epoch": 1.9857142857142858,
+      "grad_norm": 1.9651641845703125,
+      "learning_rate": 4.155555555555556e-05,
+      "loss": 0.3847,
+      "step": 2780
+    },
+    {
+      "epoch": 1.9928571428571429,
+      "grad_norm": 1.4134501218795776,
+      "learning_rate": 4.1518518518518525e-05,
+      "loss": 0.4465,
+      "step": 2790
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.9682196378707886,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.545,
+      "step": 2800
+    },
+    {
+      "epoch": 2.007142857142857,
+      "grad_norm": 2.259190559387207,
+      "learning_rate": 4.144444444444445e-05,
+      "loss": 0.3058,
+      "step": 2810
+    },
+    {
+      "epoch": 2.0142857142857142,
+      "grad_norm": 1.6861268281936646,
+      "learning_rate": 4.140740740740741e-05,
+      "loss": 0.4567,
+      "step": 2820
+    },
+    {
+      "epoch": 2.0214285714285714,
+      "grad_norm": 1.5168589353561401,
+      "learning_rate": 4.137037037037037e-05,
+      "loss": 0.5687,
+      "step": 2830
+    },
+    {
+      "epoch": 2.0285714285714285,
+      "grad_norm": 1.1756591796875,
+      "learning_rate": 4.133333333333333e-05,
+      "loss": 0.4118,
+      "step": 2840
+    },
+    {
+      "epoch": 2.0357142857142856,
+      "grad_norm": 1.4381860494613647,
+      "learning_rate": 4.12962962962963e-05,
+      "loss": 0.5473,
+      "step": 2850
+    },
+    {
+      "epoch": 2.0428571428571427,
+      "grad_norm": 1.710028052330017,
+      "learning_rate": 4.1259259259259256e-05,
+      "loss": 0.3362,
+      "step": 2860
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.3261126279830933,
+      "learning_rate": 4.1222222222222224e-05,
+      "loss": 0.4497,
+      "step": 2870
+    },
+    {
+      "epoch": 2.057142857142857,
+      "grad_norm": 1.5397872924804688,
+      "learning_rate": 4.1185185185185186e-05,
+      "loss": 0.3985,
+      "step": 2880
+    },
+    {
+      "epoch": 2.064285714285714,
+      "grad_norm": 2.1463019847869873,
+      "learning_rate": 4.1148148148148154e-05,
+      "loss": 0.3423,
+      "step": 2890
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 1.3202670812606812,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 0.3422,
+      "step": 2900
+    },
+    {
+      "epoch": 2.0785714285714287,
+      "grad_norm": 1.6832393407821655,
+      "learning_rate": 4.107407407407408e-05,
+      "loss": 0.3243,
+      "step": 2910
+    },
+    {
+      "epoch": 2.085714285714286,
+      "grad_norm": 1.7872291803359985,
+      "learning_rate": 4.103703703703704e-05,
+      "loss": 0.4452,
+      "step": 2920
+    },
+    {
+      "epoch": 2.092857142857143,
+      "grad_norm": 2.649644613265991,
+      "learning_rate": 4.1e-05,
+      "loss": 0.4125,
+      "step": 2930
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6862508058547974,
+      "learning_rate": 4.096296296296296e-05,
+      "loss": 0.3503,
+      "step": 2940
+    },
+    {
+      "epoch": 2.107142857142857,
+      "grad_norm": 1.0415781736373901,
+      "learning_rate": 4.092592592592593e-05,
+      "loss": 0.2464,
+      "step": 2950
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 1.3061981201171875,
+      "learning_rate": 4.088888888888889e-05,
+      "loss": 0.2232,
+      "step": 2960
+    },
+    {
+      "epoch": 2.1214285714285714,
+      "grad_norm": 1.0442795753479004,
+      "learning_rate": 4.0851851851851853e-05,
+      "loss": 0.3512,
+      "step": 2970
+    },
+    {
+      "epoch": 2.1285714285714286,
+      "grad_norm": 2.4381885528564453,
+      "learning_rate": 4.0814814814814815e-05,
+      "loss": 0.4678,
+      "step": 2980
+    },
+    {
+      "epoch": 2.1357142857142857,
+      "grad_norm": 1.5145925283432007,
+      "learning_rate": 4.0777777777777783e-05,
+      "loss": 0.3655,
+      "step": 2990
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 3.514470338821411,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.3538,
+      "step": 3000
+    },
+    {
+      "epoch": 2.142857142857143,
+      "eval_loss": 0.42481884360313416,
+      "eval_rouge1": 0.8943,
+      "eval_rouge2": 0.8295,
+      "eval_rougeL": 0.8911,
+      "eval_runtime": 122.0137,
+      "eval_samples_per_second": 11.474,
+      "eval_steps_per_second": 5.737,
+      "step": 3000
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.9778640270233154,
+      "learning_rate": 4.0703703703703707e-05,
+      "loss": 0.4804,
+      "step": 3010
+    },
+    {
+      "epoch": 2.157142857142857,
+      "grad_norm": 2.8497660160064697,
+      "learning_rate": 4.066666666666667e-05,
+      "loss": 0.523,
+      "step": 3020
+    },
+    {
+      "epoch": 2.164285714285714,
+      "grad_norm": 1.317818284034729,
+      "learning_rate": 4.0629629629629636e-05,
+      "loss": 0.3694,
+      "step": 3030
+    },
+    {
+      "epoch": 2.1714285714285713,
+      "grad_norm": 1.1630916595458984,
+      "learning_rate": 4.059259259259259e-05,
+      "loss": 0.3546,
+      "step": 3040
+    },
+    {
+      "epoch": 2.1785714285714284,
+      "grad_norm": 2.114527940750122,
+      "learning_rate": 4.055555555555556e-05,
+      "loss": 0.4838,
+      "step": 3050
+    },
+    {
+      "epoch": 2.185714285714286,
+      "grad_norm": 1.771263599395752,
+      "learning_rate": 4.051851851851852e-05,
+      "loss": 0.2938,
+      "step": 3060
+    },
+    {
+      "epoch": 2.192857142857143,
+      "grad_norm": 3.463986396789551,
+      "learning_rate": 4.048148148148148e-05,
+      "loss": 0.455,
+      "step": 3070
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.023069381713867,
+      "learning_rate": 4.0444444444444444e-05,
+      "loss": 0.4449,
+      "step": 3080
+    },
+    {
+      "epoch": 2.2071428571428573,
+      "grad_norm": 2.9855751991271973,
+      "learning_rate": 4.040740740740741e-05,
+      "loss": 0.5374,
+      "step": 3090
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 2.422739267349243,
+      "learning_rate": 4.0370370370370374e-05,
+      "loss": 0.4203,
+      "step": 3100
+    },
+    {
+      "epoch": 2.2214285714285715,
+      "grad_norm": 2.097543478012085,
+      "learning_rate": 4.0333333333333336e-05,
+      "loss": 0.364,
+      "step": 3110
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 2.2496302127838135,
+      "learning_rate": 4.02962962962963e-05,
+      "loss": 0.4135,
+      "step": 3120
+    },
+    {
+      "epoch": 2.2357142857142858,
+      "grad_norm": 2.3347012996673584,
+      "learning_rate": 4.0259259259259266e-05,
+      "loss": 0.4795,
+      "step": 3130
+    },
+    {
+      "epoch": 2.242857142857143,
+      "grad_norm": 1.506218433380127,
+      "learning_rate": 4.022222222222222e-05,
+      "loss": 0.5228,
+      "step": 3140
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.160443663597107,
+      "learning_rate": 4.018518518518519e-05,
+      "loss": 0.3439,
+      "step": 3150
+    },
+    {
+      "epoch": 2.257142857142857,
+      "grad_norm": 1.8678144216537476,
+      "learning_rate": 4.014814814814815e-05,
+      "loss": 0.404,
+      "step": 3160
+    },
+    {
+      "epoch": 2.2642857142857142,
+      "grad_norm": 1.1315560340881348,
+      "learning_rate": 4.011111111111111e-05,
+      "loss": 0.3111,
+      "step": 3170
+    },
+    {
+      "epoch": 2.2714285714285714,
+      "grad_norm": 1.8081461191177368,
+      "learning_rate": 4.007407407407407e-05,
+      "loss": 0.2717,
+      "step": 3180
+    },
+    {
+      "epoch": 2.2785714285714285,
+      "grad_norm": 1.6636005640029907,
+      "learning_rate": 4.003703703703704e-05,
+      "loss": 0.3382,
+      "step": 3190
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 1.3334009647369385,
+      "learning_rate": 4e-05,
+      "loss": 0.38,
+      "step": 3200
+    },
+    {
+      "epoch": 2.2928571428571427,
+      "grad_norm": 1.4873621463775635,
+      "learning_rate": 3.9962962962962965e-05,
+      "loss": 0.2979,
+      "step": 3210
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.17378568649292,
+      "learning_rate": 3.9925925925925926e-05,
+      "loss": 0.3628,
+      "step": 3220
+    },
+    {
+      "epoch": 2.307142857142857,
+      "grad_norm": 1.3241777420043945,
+      "learning_rate": 3.9888888888888895e-05,
+      "loss": 0.3119,
+      "step": 3230
+    },
+    {
+      "epoch": 2.314285714285714,
+      "grad_norm": 1.9823285341262817,
+      "learning_rate": 3.985185185185185e-05,
+      "loss": 0.4647,
+      "step": 3240
+    },
+    {
+      "epoch": 2.3214285714285716,
+      "grad_norm": 1.6918193101882935,
+      "learning_rate": 3.981481481481482e-05,
+      "loss": 0.3695,
+      "step": 3250
+    },
+    {
+      "epoch": 2.3285714285714287,
+      "grad_norm": 2.1902389526367188,
+      "learning_rate": 3.977777777777778e-05,
+      "loss": 0.2468,
+      "step": 3260
+    },
+    {
+      "epoch": 2.335714285714286,
+      "grad_norm": 1.3570506572723389,
+      "learning_rate": 3.974074074074075e-05,
+      "loss": 0.3209,
+      "step": 3270
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 1.951711654663086,
+      "learning_rate": 3.97037037037037e-05,
+      "loss": 0.5175,
+      "step": 3280
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.741243839263916,
+      "learning_rate": 3.966666666666667e-05,
+      "loss": 0.2934,
+      "step": 3290
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 1.5889472961425781,
+      "learning_rate": 3.962962962962963e-05,
+      "loss": 0.3026,
+      "step": 3300
+    },
+    {
+      "epoch": 2.3642857142857143,
+      "grad_norm": 1.4606213569641113,
+      "learning_rate": 3.9592592592592594e-05,
+      "loss": 0.4508,
+      "step": 3310
+    },
+    {
+      "epoch": 2.3714285714285714,
+      "grad_norm": 1.5021477937698364,
+      "learning_rate": 3.9555555555555556e-05,
+      "loss": 0.2589,
+      "step": 3320
+    },
+    {
+      "epoch": 2.3785714285714286,
+      "grad_norm": 1.8877885341644287,
+      "learning_rate": 3.9518518518518524e-05,
+      "loss": 0.4115,
+      "step": 3330
+    },
+    {
+      "epoch": 2.3857142857142857,
+      "grad_norm": 1.809822678565979,
+      "learning_rate": 3.9481481481481485e-05,
+      "loss": 0.3844,
+      "step": 3340
+    },
+    {
+      "epoch": 2.392857142857143,
+      "grad_norm": 1.2999638319015503,
+      "learning_rate": 3.944444444444445e-05,
+      "loss": 0.4028,
+      "step": 3350
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4639837741851807,
+      "learning_rate": 3.940740740740741e-05,
+      "loss": 0.3551,
+      "step": 3360
+    },
+    {
+      "epoch": 2.407142857142857,
+      "grad_norm": 1.1001754999160767,
+      "learning_rate": 3.937037037037038e-05,
+      "loss": 0.4001,
+      "step": 3370
+    },
+    {
+      "epoch": 2.414285714285714,
+      "grad_norm": 2.272892713546753,
+      "learning_rate": 3.933333333333333e-05,
+      "loss": 0.3048,
+      "step": 3380
+    },
+    {
+      "epoch": 2.4214285714285713,
+      "grad_norm": 2.085908889770508,
+      "learning_rate": 3.92962962962963e-05,
+      "loss": 0.5788,
+      "step": 3390
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 1.317700743675232,
+      "learning_rate": 3.925925925925926e-05,
+      "loss": 0.2922,
+      "step": 3400
+    },
+    {
+      "epoch": 2.435714285714286,
+      "grad_norm": 2.372558832168579,
+      "learning_rate": 3.922222222222223e-05,
+      "loss": 0.4581,
+      "step": 3410
+    },
+    {
+      "epoch": 2.442857142857143,
+      "grad_norm": 1.3307292461395264,
+      "learning_rate": 3.9185185185185185e-05,
+      "loss": 0.4553,
+      "step": 3420
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.9228068590164185,
+      "learning_rate": 3.914814814814815e-05,
+      "loss": 0.497,
+      "step": 3430
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 1.071590542793274,
+      "learning_rate": 3.9111111111111115e-05,
+      "loss": 0.4532,
+      "step": 3440
+    },
+    {
+      "epoch": 2.4642857142857144,
+      "grad_norm": 1.9603391885757446,
+      "learning_rate": 3.9074074074074076e-05,
+      "loss": 0.3808,
+      "step": 3450
+    },
+    {
+      "epoch": 2.4714285714285715,
+      "grad_norm": 1.2152074575424194,
+      "learning_rate": 3.903703703703704e-05,
+      "loss": 0.3011,
+      "step": 3460
+    },
+    {
+      "epoch": 2.4785714285714286,
+      "grad_norm": 1.532478928565979,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.3403,
+      "step": 3470
+    },
+    {
+      "epoch": 2.4857142857142858,
+      "grad_norm": 1.5086220502853394,
+      "learning_rate": 3.896296296296296e-05,
+      "loss": 0.4835,
+      "step": 3480
+    },
+    {
+      "epoch": 2.492857142857143,
+      "grad_norm": 1.0601118803024292,
+      "learning_rate": 3.892592592592593e-05,
+      "loss": 0.4057,
+      "step": 3490
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7907903790473938,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.3183,
+      "step": 3500
+    },
+    {
+      "epoch": 2.507142857142857,
+      "grad_norm": 1.8523814678192139,
+      "learning_rate": 3.885185185185186e-05,
+      "loss": 0.4329,
+      "step": 3510
+    },
+    {
+      "epoch": 2.5142857142857142,
+      "grad_norm": 1.9627383947372437,
+      "learning_rate": 3.8814814814814814e-05,
+      "loss": 0.3041,
+      "step": 3520
+    },
+    {
+      "epoch": 2.5214285714285714,
+      "grad_norm": 0.6192536354064941,
+      "learning_rate": 3.877777777777778e-05,
+      "loss": 0.3271,
+      "step": 3530
+    },
+    {
+      "epoch": 2.5285714285714285,
+      "grad_norm": 1.3901042938232422,
+      "learning_rate": 3.8740740740740744e-05,
+      "loss": 0.2562,
+      "step": 3540
+    },
+    {
+      "epoch": 2.5357142857142856,
+      "grad_norm": 1.991752028465271,
+      "learning_rate": 3.8703703703703705e-05,
+      "loss": 0.4283,
+      "step": 3550
+    },
+    {
+      "epoch": 2.5428571428571427,
+      "grad_norm": 1.219382882118225,
+      "learning_rate": 3.866666666666667e-05,
+      "loss": 0.4232,
+      "step": 3560
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.72744083404541,
+      "learning_rate": 3.8629629629629635e-05,
+      "loss": 0.3482,
+      "step": 3570
+    },
+    {
+      "epoch": 2.557142857142857,
+      "grad_norm": 1.6782621145248413,
+      "learning_rate": 3.85925925925926e-05,
+      "loss": 0.3302,
+      "step": 3580
+    },
+    {
+      "epoch": 2.564285714285714,
+      "grad_norm": 1.0238265991210938,
+      "learning_rate": 3.855555555555556e-05,
+      "loss": 0.458,
+      "step": 3590
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 2.212013006210327,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.4127,
+      "step": 3600
+    },
+    {
+      "epoch": 2.5785714285714287,
+      "grad_norm": 1.5761399269104004,
+      "learning_rate": 3.848148148148149e-05,
+      "loss": 0.402,
+      "step": 3610
+    },
+    {
+      "epoch": 2.585714285714286,
+      "grad_norm": 1.2036465406417847,
+      "learning_rate": 3.844444444444444e-05,
+      "loss": 0.5576,
+      "step": 3620
+    },
+    {
+      "epoch": 2.592857142857143,
+      "grad_norm": 1.8674002885818481,
+      "learning_rate": 3.840740740740741e-05,
+      "loss": 0.364,
+      "step": 3630
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.466834545135498,
+      "learning_rate": 3.837037037037037e-05,
+      "loss": 0.3523,
+      "step": 3640
+    },
+    {
+      "epoch": 2.607142857142857,
+      "grad_norm": 1.57899010181427,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 0.4046,
+      "step": 3650
+    },
+    {
+      "epoch": 2.6142857142857143,
+      "grad_norm": 0.9730345010757446,
+      "learning_rate": 3.8296296296296296e-05,
+      "loss": 0.3132,
+      "step": 3660
+    },
+    {
+      "epoch": 2.6214285714285714,
+      "grad_norm": 1.3017544746398926,
+      "learning_rate": 3.8259259259259264e-05,
+      "loss": 0.3023,
+      "step": 3670
+    },
+    {
+      "epoch": 2.6285714285714286,
+      "grad_norm": 1.6368205547332764,
+      "learning_rate": 3.8222222222222226e-05,
+      "loss": 0.4363,
+      "step": 3680
+    },
+    {
+      "epoch": 2.6357142857142857,
+      "grad_norm": 1.2852121591567993,
+      "learning_rate": 3.818518518518519e-05,
+      "loss": 0.2896,
+      "step": 3690
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 3.6991841793060303,
+      "learning_rate": 3.814814814814815e-05,
+      "loss": 0.353,
+      "step": 3700
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.70285701751709,
+      "learning_rate": 3.811111111111112e-05,
+      "loss": 0.4217,
+      "step": 3710
+    },
+    {
+      "epoch": 2.657142857142857,
+      "grad_norm": 1.140811800956726,
+      "learning_rate": 3.807407407407408e-05,
+      "loss": 0.3253,
+      "step": 3720
+    },
+    {
+      "epoch": 2.664285714285714,
+      "grad_norm": 1.2905789613723755,
+      "learning_rate": 3.803703703703704e-05,
+      "loss": 0.3051,
+      "step": 3730
+    },
+    {
+      "epoch": 2.6714285714285713,
+      "grad_norm": 1.4326887130737305,
+      "learning_rate": 3.8e-05,
+      "loss": 0.3999,
+      "step": 3740
+    },
+    {
+      "epoch": 2.678571428571429,
+      "grad_norm": 1.1789475679397583,
+      "learning_rate": 3.7962962962962964e-05,
+      "loss": 0.4631,
+      "step": 3750
+    },
+    {
+      "epoch": 2.685714285714286,
+      "grad_norm": 2.0444328784942627,
+      "learning_rate": 3.7925925925925925e-05,
+      "loss": 0.4449,
+      "step": 3760
+    },
+    {
+      "epoch": 2.692857142857143,
+      "grad_norm": 2.1025991439819336,
+      "learning_rate": 3.7888888888888894e-05,
+      "loss": 0.3513,
+      "step": 3770
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.8492026329040527,
+      "learning_rate": 3.7851851851851855e-05,
+      "loss": 0.4006,
+      "step": 3780
+    },
+    {
+      "epoch": 2.7071428571428573,
+      "grad_norm": 1.3439162969589233,
+      "learning_rate": 3.781481481481482e-05,
+      "loss": 0.2806,
+      "step": 3790
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 1.4200384616851807,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.3759,
+      "step": 3800
+    },
+    {
+      "epoch": 2.7214285714285715,
+      "grad_norm": 1.9567861557006836,
+      "learning_rate": 3.774074074074074e-05,
+      "loss": 0.1772,
+      "step": 3810
+    },
+    {
+      "epoch": 2.7285714285714286,
+      "grad_norm": 1.3466306924819946,
+      "learning_rate": 3.770370370370371e-05,
+      "loss": 0.399,
+      "step": 3820
+    },
+    {
+      "epoch": 2.7357142857142858,
+      "grad_norm": 1.6046024560928345,
+      "learning_rate": 3.766666666666667e-05,
+      "loss": 0.347,
+      "step": 3830
+    },
+    {
+      "epoch": 2.742857142857143,
+      "grad_norm": 2.190568447113037,
+      "learning_rate": 3.762962962962963e-05,
+      "loss": 0.3977,
+      "step": 3840
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.7715718746185303,
+      "learning_rate": 3.759259259259259e-05,
+      "loss": 0.4385,
+      "step": 3850
+    },
+    {
+      "epoch": 2.757142857142857,
+      "grad_norm": 3.19500994682312,
+      "learning_rate": 3.7555555555555554e-05,
+      "loss": 0.3631,
+      "step": 3860
+    },
+    {
+      "epoch": 2.7642857142857142,
+      "grad_norm": 2.2222607135772705,
+      "learning_rate": 3.751851851851852e-05,
+      "loss": 0.3565,
+      "step": 3870
+    },
+    {
+      "epoch": 2.7714285714285714,
+      "grad_norm": 1.9959403276443481,
+      "learning_rate": 3.7481481481481484e-05,
+      "loss": 0.3629,
+      "step": 3880
+    },
+    {
+      "epoch": 2.7785714285714285,
+      "grad_norm": 1.3207546472549438,
+      "learning_rate": 3.7444444444444446e-05,
+      "loss": 0.2911,
+      "step": 3890
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 2.0290961265563965,
+      "learning_rate": 3.740740740740741e-05,
+      "loss": 0.3072,
+      "step": 3900
+    },
+    {
+      "epoch": 2.7928571428571427,
+      "grad_norm": 1.3728725910186768,
+      "learning_rate": 3.737037037037037e-05,
+      "loss": 0.3858,
+      "step": 3910
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.541598320007324,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 0.3487,
+      "step": 3920
+    },
+    {
+      "epoch": 2.807142857142857,
+      "grad_norm": 2.3327584266662598,
+      "learning_rate": 3.72962962962963e-05,
+      "loss": 0.3535,
+      "step": 3930
+    },
+    {
+      "epoch": 2.814285714285714,
+      "grad_norm": 2.546766757965088,
+      "learning_rate": 3.725925925925926e-05,
+      "loss": 0.3462,
+      "step": 3940
+    },
+    {
+      "epoch": 2.821428571428571,
+      "grad_norm": 2.351959705352783,
+      "learning_rate": 3.722222222222222e-05,
+      "loss": 0.2781,
+      "step": 3950
+    },
+    {
+      "epoch": 2.8285714285714287,
+      "grad_norm": 1.9349900484085083,
+      "learning_rate": 3.718518518518519e-05,
+      "loss": 0.2442,
+      "step": 3960
+    },
+    {
+      "epoch": 2.835714285714286,
+      "grad_norm": 2.2020022869110107,
+      "learning_rate": 3.714814814814815e-05,
+      "loss": 0.3396,
+      "step": 3970
+    },
+    {
+      "epoch": 2.842857142857143,
+      "grad_norm": 1.5161465406417847,
+      "learning_rate": 3.7111111111111113e-05,
+      "loss": 0.3722,
+      "step": 3980
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.7403453588485718,
+      "learning_rate": 3.7074074074074075e-05,
+      "loss": 0.4227,
+      "step": 3990
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 1.9142546653747559,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.3259,
+      "step": 4000
+    },
+    {
+      "epoch": 2.857142857142857,
+      "eval_loss": 0.38673722743988037,
+      "eval_rouge1": 0.8974,
+      "eval_rouge2": 0.8331,
+      "eval_rougeL": 0.8942,
+      "eval_runtime": 122.1383,
+      "eval_samples_per_second": 11.462,
+      "eval_steps_per_second": 5.731,
+      "step": 4000
+    },
+    {
+      "epoch": 2.8642857142857143,
+      "grad_norm": 1.5975255966186523,
+      "learning_rate": 3.7e-05,
+      "loss": 0.3732,
+      "step": 4010
+    },
+    {
+      "epoch": 2.8714285714285714,
+      "grad_norm": 1.4830248355865479,
+      "learning_rate": 3.6962962962962966e-05,
+      "loss": 0.5093,
+      "step": 4020
+    },
+    {
+      "epoch": 2.8785714285714286,
+      "grad_norm": 2.504650354385376,
+      "learning_rate": 3.692592592592593e-05,
+      "loss": 0.3302,
+      "step": 4030
+    },
+    {
+      "epoch": 2.8857142857142857,
+      "grad_norm": 2.349452495574951,
+      "learning_rate": 3.688888888888889e-05,
+      "loss": 0.3596,
+      "step": 4040
+    },
+    {
+      "epoch": 2.892857142857143,
+      "grad_norm": 1.398964762687683,
+      "learning_rate": 3.685185185185185e-05,
+      "loss": 0.3494,
+      "step": 4050
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.212738513946533,
+      "learning_rate": 3.681481481481482e-05,
+      "loss": 0.3691,
+      "step": 4060
+    },
+    {
+      "epoch": 2.907142857142857,
+      "grad_norm": 2.20845627784729,
+      "learning_rate": 3.677777777777778e-05,
+      "loss": 0.2974,
+      "step": 4070
+    },
+    {
+      "epoch": 2.914285714285714,
+      "grad_norm": 1.2226334810256958,
+      "learning_rate": 3.674074074074074e-05,
+      "loss": 0.3173,
+      "step": 4080
+    },
+    {
+      "epoch": 2.9214285714285713,
+      "grad_norm": 2.2203428745269775,
+      "learning_rate": 3.6703703703703704e-05,
+      "loss": 0.4473,
+      "step": 4090
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 1.487853765487671,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.2653,
+      "step": 4100
+    },
+    {
+      "epoch": 2.935714285714286,
+      "grad_norm": 1.6347614526748657,
+      "learning_rate": 3.662962962962963e-05,
+      "loss": 0.3563,
+      "step": 4110
+    },
+    {
+      "epoch": 2.942857142857143,
+      "grad_norm": 2.2722184658050537,
+      "learning_rate": 3.6592592592592596e-05,
+      "loss": 0.4975,
+      "step": 4120
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.747530460357666,
+      "learning_rate": 3.655555555555556e-05,
+      "loss": 0.2357,
+      "step": 4130
+    },
+    {
+      "epoch": 2.9571428571428573,
+      "grad_norm": 1.628596544265747,
+      "learning_rate": 3.651851851851852e-05,
+      "loss": 0.3674,
+      "step": 4140
+    },
+    {
+      "epoch": 2.9642857142857144,
+      "grad_norm": 1.0486435890197754,
+      "learning_rate": 3.648148148148148e-05,
+      "loss": 0.3314,
+      "step": 4150
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "grad_norm": 2.523879289627075,
+      "learning_rate": 3.644444444444445e-05,
+      "loss": 0.4421,
+      "step": 4160
+    },
+    {
+      "epoch": 2.9785714285714286,
+      "grad_norm": 1.4641958475112915,
+      "learning_rate": 3.6407407407407403e-05,
+      "loss": 0.4135,
+      "step": 4170
+    },
+    {
+      "epoch": 2.9857142857142858,
+      "grad_norm": 2.672769784927368,
+      "learning_rate": 3.637037037037037e-05,
+      "loss": 0.3527,
+      "step": 4180
+    },
+    {
+      "epoch": 2.992857142857143,
+      "grad_norm": 0.5795308351516724,
+      "learning_rate": 3.633333333333333e-05,
+      "loss": 0.2326,
+      "step": 4190
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.873579978942871,
+      "learning_rate": 3.62962962962963e-05,
+      "loss": 0.3679,
+      "step": 4200
+    },
+    {
+      "epoch": 3.007142857142857,
+      "grad_norm": 1.7640775442123413,
+      "learning_rate": 3.6259259259259256e-05,
+      "loss": 0.4778,
+      "step": 4210
+    },
+    {
+      "epoch": 3.0142857142857142,
+      "grad_norm": 1.9458075761795044,
+      "learning_rate": 3.6222222222222225e-05,
+      "loss": 0.4054,
+      "step": 4220
+    },
+    {
+      "epoch": 3.0214285714285714,
+      "grad_norm": 1.1568126678466797,
+      "learning_rate": 3.6185185185185186e-05,
+      "loss": 0.2249,
+      "step": 4230
+    },
+    {
+      "epoch": 3.0285714285714285,
+      "grad_norm": 1.3655381202697754,
+      "learning_rate": 3.614814814814815e-05,
+      "loss": 0.3993,
+      "step": 4240
+    },
+    {
+      "epoch": 3.0357142857142856,
+      "grad_norm": 2.0403196811676025,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 0.3366,
+      "step": 4250
+    },
+    {
+      "epoch": 3.0428571428571427,
+      "grad_norm": 1.9888697862625122,
+      "learning_rate": 3.607407407407408e-05,
+      "loss": 0.3033,
+      "step": 4260
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.3648616075515747,
+      "learning_rate": 3.603703703703704e-05,
+      "loss": 0.2874,
+      "step": 4270
+    },
+    {
+      "epoch": 3.057142857142857,
+      "grad_norm": 2.602613925933838,
+      "learning_rate": 3.6e-05,
+      "loss": 0.4086,
+      "step": 4280
+    },
+    {
+      "epoch": 3.064285714285714,
+      "grad_norm": 2.5918185710906982,
+      "learning_rate": 3.596296296296296e-05,
+      "loss": 0.393,
+      "step": 4290
+    },
+    {
+      "epoch": 3.0714285714285716,
+      "grad_norm": 1.8195433616638184,
+      "learning_rate": 3.592592592592593e-05,
+      "loss": 0.3361,
+      "step": 4300
+    },
+    {
+      "epoch": 3.0785714285714287,
+      "grad_norm": 1.8855136632919312,
+      "learning_rate": 3.5888888888888886e-05,
+      "loss": 0.3205,
+      "step": 4310
+    },
+    {
+      "epoch": 3.085714285714286,
+      "grad_norm": 2.7412662506103516,
+      "learning_rate": 3.5851851851851854e-05,
+      "loss": 0.2659,
+      "step": 4320
+    },
+    {
+      "epoch": 3.092857142857143,
+      "grad_norm": 1.880436658859253,
+      "learning_rate": 3.5814814814814815e-05,
+      "loss": 0.49,
+      "step": 4330
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.6828274726867676,
+      "learning_rate": 3.577777777777778e-05,
+      "loss": 0.2933,
+      "step": 4340
+    },
+    {
+      "epoch": 3.107142857142857,
+      "grad_norm": 1.0517287254333496,
+      "learning_rate": 3.574074074074074e-05,
+      "loss": 0.3563,
+      "step": 4350
+    },
+    {
+      "epoch": 3.1142857142857143,
+      "grad_norm": 1.3242154121398926,
+      "learning_rate": 3.570370370370371e-05,
+      "loss": 0.3765,
+      "step": 4360
+    },
+    {
+      "epoch": 3.1214285714285714,
+      "grad_norm": 2.0899312496185303,
+      "learning_rate": 3.566666666666667e-05,
+      "loss": 0.3664,
+      "step": 4370
+    },
+    {
+      "epoch": 3.1285714285714286,
+      "grad_norm": 2.0286014080047607,
+      "learning_rate": 3.562962962962963e-05,
+      "loss": 0.2622,
+      "step": 4380
+    },
+    {
+      "epoch": 3.1357142857142857,
+      "grad_norm": 2.5074400901794434,
+      "learning_rate": 3.559259259259259e-05,
+      "loss": 0.321,
+      "step": 4390
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 1.4080287218093872,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 0.4035,
+      "step": 4400
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.923890471458435,
+      "learning_rate": 3.5518518518518515e-05,
+      "loss": 0.2775,
+      "step": 4410
+    },
+    {
+      "epoch": 3.157142857142857,
+      "grad_norm": 0.806591272354126,
+      "learning_rate": 3.548148148148148e-05,
+      "loss": 0.3149,
+      "step": 4420
+    },
+    {
+      "epoch": 3.164285714285714,
+      "grad_norm": 2.197736978530884,
+      "learning_rate": 3.5444444444444445e-05,
+      "loss": 0.4368,
+      "step": 4430
+    },
+    {
+      "epoch": 3.1714285714285713,
+      "grad_norm": 1.6943881511688232,
+      "learning_rate": 3.540740740740741e-05,
+      "loss": 0.2793,
+      "step": 4440
+    },
+    {
+      "epoch": 3.1785714285714284,
+      "grad_norm": 2.5460283756256104,
+      "learning_rate": 3.537037037037037e-05,
+      "loss": 0.4057,
+      "step": 4450
+    },
+    {
+      "epoch": 3.185714285714286,
+      "grad_norm": 1.579908013343811,
+      "learning_rate": 3.5333333333333336e-05,
+      "loss": 0.3016,
+      "step": 4460
+    },
+    {
+      "epoch": 3.192857142857143,
+      "grad_norm": 1.9137247800827026,
+      "learning_rate": 3.52962962962963e-05,
+      "loss": 0.3437,
+      "step": 4470
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.510328769683838,
+      "learning_rate": 3.525925925925926e-05,
+      "loss": 0.585,
+      "step": 4480
+    },
+    {
+      "epoch": 3.2071428571428573,
+      "grad_norm": 0.9775506854057312,
+      "learning_rate": 3.522222222222222e-05,
+      "loss": 0.2651,
+      "step": 4490
+    },
+    {
+      "epoch": 3.2142857142857144,
+      "grad_norm": 1.7614684104919434,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.3089,
+      "step": 4500
+    },
+    {
+      "epoch": 3.2214285714285715,
+      "grad_norm": 1.9103621244430542,
+      "learning_rate": 3.514814814814815e-05,
+      "loss": 0.342,
+      "step": 4510
+    },
+    {
+      "epoch": 3.2285714285714286,
+      "grad_norm": 1.4587639570236206,
+      "learning_rate": 3.511111111111111e-05,
+      "loss": 0.2592,
+      "step": 4520
+    },
+    {
+      "epoch": 3.2357142857142858,
+      "grad_norm": 1.3419288396835327,
+      "learning_rate": 3.5074074074074074e-05,
+      "loss": 0.4185,
+      "step": 4530
+    },
+    {
+      "epoch": 3.242857142857143,
+      "grad_norm": 1.6199047565460205,
+      "learning_rate": 3.503703703703704e-05,
+      "loss": 0.256,
+      "step": 4540
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.230350136756897,
+      "learning_rate": 3.5e-05,
+      "loss": 0.3304,
+      "step": 4550
+    },
+    {
+      "epoch": 3.257142857142857,
+      "grad_norm": 3.087888240814209,
+      "learning_rate": 3.4962962962962965e-05,
+      "loss": 0.3351,
+      "step": 4560
+    },
+    {
+      "epoch": 3.2642857142857142,
+      "grad_norm": 1.4498260021209717,
+      "learning_rate": 3.492592592592593e-05,
+      "loss": 0.2753,
+      "step": 4570
+    },
+    {
+      "epoch": 3.2714285714285714,
+      "grad_norm": 1.1032336950302124,
+      "learning_rate": 3.4888888888888895e-05,
+      "loss": 0.3709,
+      "step": 4580
+    },
+    {
+      "epoch": 3.2785714285714285,
+      "grad_norm": 1.5177497863769531,
+      "learning_rate": 3.485185185185185e-05,
+      "loss": 0.276,
+      "step": 4590
+    },
+    {
+      "epoch": 3.2857142857142856,
+      "grad_norm": 1.2596136331558228,
+      "learning_rate": 3.481481481481482e-05,
+      "loss": 0.3482,
+      "step": 4600
+    },
+    {
+      "epoch": 3.2928571428571427,
+      "grad_norm": 1.9895663261413574,
+      "learning_rate": 3.477777777777778e-05,
+      "loss": 0.3738,
+      "step": 4610
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.2930881977081299,
+      "learning_rate": 3.474074074074074e-05,
+      "loss": 0.4263,
+      "step": 4620
+    },
+    {
+      "epoch": 3.307142857142857,
+      "grad_norm": 2.276385545730591,
+      "learning_rate": 3.47037037037037e-05,
+      "loss": 0.2267,
+      "step": 4630
+    },
+    {
+      "epoch": 3.314285714285714,
+      "grad_norm": 0.9766007661819458,
+      "learning_rate": 3.466666666666667e-05,
+      "loss": 0.2217,
+      "step": 4640
+    },
+    {
+      "epoch": 3.3214285714285716,
+      "grad_norm": 1.5184674263000488,
+      "learning_rate": 3.4629629629629626e-05,
+      "loss": 0.2788,
+      "step": 4650
+    },
+    {
+      "epoch": 3.3285714285714287,
+      "grad_norm": 1.5145732164382935,
+      "learning_rate": 3.4592592592592594e-05,
+      "loss": 0.3291,
+      "step": 4660
+    },
+    {
+      "epoch": 3.335714285714286,
+      "grad_norm": 1.4273874759674072,
+      "learning_rate": 3.4555555555555556e-05,
+      "loss": 0.2854,
+      "step": 4670
+    },
+    {
+      "epoch": 3.342857142857143,
+      "grad_norm": 2.783701181411743,
+      "learning_rate": 3.4518518518518524e-05,
+      "loss": 0.3518,
+      "step": 4680
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.3359688520431519,
+      "learning_rate": 3.448148148148148e-05,
+      "loss": 0.2239,
+      "step": 4690
+    },
+    {
+      "epoch": 3.357142857142857,
+      "grad_norm": 2.246824264526367,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.3206,
+      "step": 4700
+    },
+    {
+      "epoch": 3.3642857142857143,
+      "grad_norm": 1.7839916944503784,
+      "learning_rate": 3.440740740740741e-05,
+      "loss": 0.3189,
+      "step": 4710
+    },
+    {
+      "epoch": 3.3714285714285714,
+      "grad_norm": 1.0196881294250488,
+      "learning_rate": 3.437037037037037e-05,
+      "loss": 0.2318,
+      "step": 4720
+    },
+    {
+      "epoch": 3.3785714285714286,
+      "grad_norm": 2.228317975997925,
+      "learning_rate": 3.433333333333333e-05,
+      "loss": 0.4033,
+      "step": 4730
+    },
+    {
+      "epoch": 3.3857142857142857,
+      "grad_norm": 2.0231473445892334,
+      "learning_rate": 3.42962962962963e-05,
+      "loss": 0.3854,
+      "step": 4740
+    },
+    {
+      "epoch": 3.392857142857143,
+      "grad_norm": 2.074925422668457,
+      "learning_rate": 3.425925925925926e-05,
+      "loss": 0.3778,
+      "step": 4750
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.2508392333984375,
+      "learning_rate": 3.4222222222222224e-05,
+      "loss": 0.3299,
+      "step": 4760
+    },
+    {
+      "epoch": 3.407142857142857,
+      "grad_norm": 1.0920076370239258,
+      "learning_rate": 3.4185185185185185e-05,
+      "loss": 0.3798,
+      "step": 4770
+    },
+    {
+      "epoch": 3.414285714285714,
+      "grad_norm": 1.8113828897476196,
+      "learning_rate": 3.4148148148148153e-05,
+      "loss": 0.2903,
+      "step": 4780
+    },
+    {
+      "epoch": 3.4214285714285713,
+      "grad_norm": 1.6218737363815308,
+      "learning_rate": 3.411111111111111e-05,
+      "loss": 0.2593,
+      "step": 4790
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 1.0635234117507935,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.4388,
+      "step": 4800
+    },
+    {
+      "epoch": 3.435714285714286,
+      "grad_norm": 2.585700273513794,
+      "learning_rate": 3.403703703703704e-05,
+      "loss": 0.3368,
+      "step": 4810
+    },
+    {
+      "epoch": 3.442857142857143,
+      "grad_norm": 1.0704694986343384,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.2196,
+      "step": 4820
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.3177589178085327,
+      "learning_rate": 3.396296296296296e-05,
+      "loss": 0.3104,
+      "step": 4830
+    },
+    {
+      "epoch": 3.4571428571428573,
+      "grad_norm": 1.834241271018982,
+      "learning_rate": 3.392592592592593e-05,
+      "loss": 0.3413,
+      "step": 4840
+    },
+    {
+      "epoch": 3.4642857142857144,
+      "grad_norm": 1.8859339952468872,
+      "learning_rate": 3.388888888888889e-05,
+      "loss": 0.2593,
+      "step": 4850
+    },
+    {
+      "epoch": 3.4714285714285715,
+      "grad_norm": 1.452728271484375,
+      "learning_rate": 3.385185185185185e-05,
+      "loss": 0.3029,
+      "step": 4860
+    },
+    {
+      "epoch": 3.4785714285714286,
+      "grad_norm": 2.170774221420288,
+      "learning_rate": 3.3814814814814814e-05,
+      "loss": 0.3372,
+      "step": 4870
+    },
+    {
+      "epoch": 3.4857142857142858,
+      "grad_norm": 1.8695834875106812,
+      "learning_rate": 3.377777777777778e-05,
+      "loss": 0.3428,
+      "step": 4880
+    },
+    {
+      "epoch": 3.492857142857143,
+      "grad_norm": 1.74647855758667,
+      "learning_rate": 3.3740740740740744e-05,
+      "loss": 0.3351,
+      "step": 4890
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.3349127769470215,
+      "learning_rate": 3.3703703703703706e-05,
+      "loss": 0.2733,
+      "step": 4900
+    },
+    {
+      "epoch": 3.507142857142857,
+      "grad_norm": 2.73463773727417,
+      "learning_rate": 3.366666666666667e-05,
+      "loss": 0.2979,
+      "step": 4910
+    },
+    {
+      "epoch": 3.5142857142857142,
+      "grad_norm": 1.3546210527420044,
+      "learning_rate": 3.3629629629629636e-05,
+      "loss": 0.3521,
+      "step": 4920
+    },
+    {
+      "epoch": 3.5214285714285714,
+      "grad_norm": 1.617336630821228,
+      "learning_rate": 3.359259259259259e-05,
+      "loss": 0.2758,
+      "step": 4930
+    },
+    {
+      "epoch": 3.5285714285714285,
+      "grad_norm": 2.998967409133911,
+      "learning_rate": 3.355555555555556e-05,
+      "loss": 0.4193,
+      "step": 4940
+    },
+    {
+      "epoch": 3.5357142857142856,
+      "grad_norm": 1.8004390001296997,
+      "learning_rate": 3.351851851851852e-05,
+      "loss": 0.3936,
+      "step": 4950
+    },
+    {
+      "epoch": 3.5428571428571427,
+      "grad_norm": 1.4228971004486084,
+      "learning_rate": 3.348148148148148e-05,
+      "loss": 0.3563,
+      "step": 4960
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.5617480278015137,
+      "learning_rate": 3.3444444444444443e-05,
+      "loss": 0.2492,
+      "step": 4970
+    },
+    {
+      "epoch": 3.557142857142857,
+      "grad_norm": 1.3880919218063354,
+      "learning_rate": 3.340740740740741e-05,
+      "loss": 0.1791,
+      "step": 4980
+    },
+    {
+      "epoch": 3.564285714285714,
+      "grad_norm": 2.3505630493164062,
+      "learning_rate": 3.337037037037037e-05,
+      "loss": 0.4009,
+      "step": 4990
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.9086794853210449,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.2826,
+      "step": 5000
+    },
+    {
+      "epoch": 3.571428571428571,
+      "eval_loss": 0.3789908289909363,
+      "eval_rouge1": 0.8999,
+      "eval_rouge2": 0.8372,
+      "eval_rougeL": 0.8969,
+      "eval_runtime": 122.23,
+      "eval_samples_per_second": 11.454,
+      "eval_steps_per_second": 5.727,
+      "step": 5000
+    },
+    {
+      "epoch": 3.5785714285714287,
+      "grad_norm": 1.0208678245544434,
+      "learning_rate": 3.3296296296296296e-05,
+      "loss": 0.295,
+      "step": 5010
+    },
+    {
+      "epoch": 3.585714285714286,
+      "grad_norm": 3.03141713142395,
+      "learning_rate": 3.3259259259259265e-05,
+      "loss": 0.3813,
+      "step": 5020
+    },
+    {
+      "epoch": 3.592857142857143,
+      "grad_norm": 1.7845333814620972,
+      "learning_rate": 3.322222222222222e-05,
+      "loss": 0.2526,
+      "step": 5030
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 4.314096450805664,
+      "learning_rate": 3.318518518518519e-05,
+      "loss": 0.3498,
+      "step": 5040
+    },
+    {
+      "epoch": 3.607142857142857,
+      "grad_norm": 1.5270274877548218,
+      "learning_rate": 3.314814814814815e-05,
+      "loss": 0.3204,
+      "step": 5050
+    },
+    {
+      "epoch": 3.6142857142857143,
+      "grad_norm": 2.036738157272339,
+      "learning_rate": 3.311111111111112e-05,
+      "loss": 0.3416,
+      "step": 5060
+    },
+    {
+      "epoch": 3.6214285714285714,
+      "grad_norm": 2.2504570484161377,
+      "learning_rate": 3.307407407407407e-05,
+      "loss": 0.3781,
+      "step": 5070
+    },
+    {
+      "epoch": 3.6285714285714286,
+      "grad_norm": 1.749518632888794,
+      "learning_rate": 3.303703703703704e-05,
+      "loss": 0.2299,
+      "step": 5080
+    },
+    {
+      "epoch": 3.6357142857142857,
+      "grad_norm": 2.1878907680511475,
+      "learning_rate": 3.3e-05,
+      "loss": 0.3692,
+      "step": 5090
+    },
+    {
+      "epoch": 3.642857142857143,
+      "grad_norm": 1.829394817352295,
+      "learning_rate": 3.2962962962962964e-05,
+      "loss": 0.3095,
+      "step": 5100
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 2.5994794368743896,
+      "learning_rate": 3.2925925925925926e-05,
+      "loss": 0.431,
+      "step": 5110
+    },
+    {
+      "epoch": 3.657142857142857,
+      "grad_norm": 1.2319742441177368,
+      "learning_rate": 3.2888888888888894e-05,
+      "loss": 0.336,
+      "step": 5120
+    },
+    {
+      "epoch": 3.664285714285714,
+      "grad_norm": 2.169063091278076,
+      "learning_rate": 3.2851851851851856e-05,
+      "loss": 0.293,
+      "step": 5130
+    },
+    {
+      "epoch": 3.6714285714285713,
+      "grad_norm": 1.7120137214660645,
+      "learning_rate": 3.281481481481482e-05,
+      "loss": 0.3439,
+      "step": 5140
+    },
+    {
+      "epoch": 3.678571428571429,
+      "grad_norm": 1.5415689945220947,
+      "learning_rate": 3.277777777777778e-05,
+      "loss": 0.3912,
+      "step": 5150
+    },
+    {
+      "epoch": 3.685714285714286,
+      "grad_norm": 2.2880282402038574,
+      "learning_rate": 3.274074074074075e-05,
+      "loss": 0.2352,
+      "step": 5160
+    },
+    {
+      "epoch": 3.692857142857143,
+      "grad_norm": 1.7133980989456177,
+      "learning_rate": 3.27037037037037e-05,
+      "loss": 0.5397,
+      "step": 5170
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.9661128520965576,
+      "learning_rate": 3.266666666666667e-05,
+      "loss": 0.4496,
+      "step": 5180
+    },
+    {
+      "epoch": 3.7071428571428573,
+      "grad_norm": 1.444551944732666,
+      "learning_rate": 3.262962962962963e-05,
+      "loss": 0.3201,
+      "step": 5190
+    },
+    {
+      "epoch": 3.7142857142857144,
+      "grad_norm": 1.7919954061508179,
+      "learning_rate": 3.25925925925926e-05,
+      "loss": 0.3721,
+      "step": 5200
+    },
+    {
+      "epoch": 3.7214285714285715,
+      "grad_norm": 2.4862735271453857,
+      "learning_rate": 3.2555555555555555e-05,
+      "loss": 0.2511,
+      "step": 5210
+    },
+    {
+      "epoch": 3.7285714285714286,
+      "grad_norm": 1.0694047212600708,
+      "learning_rate": 3.251851851851852e-05,
+      "loss": 0.1418,
+      "step": 5220
+    },
+    {
+      "epoch": 3.7357142857142858,
+      "grad_norm": 2.4438931941986084,
+      "learning_rate": 3.2481481481481485e-05,
+      "loss": 0.2473,
+      "step": 5230
+    },
+    {
+      "epoch": 3.742857142857143,
+      "grad_norm": 1.9673523902893066,
+      "learning_rate": 3.2444444444444446e-05,
+      "loss": 0.3251,
+      "step": 5240
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 2.5299620628356934,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 0.3862,
+      "step": 5250
+    },
+    {
+      "epoch": 3.757142857142857,
+      "grad_norm": 1.1709238290786743,
+      "learning_rate": 3.2370370370370376e-05,
+      "loss": 0.3156,
+      "step": 5260
+    },
+    {
+      "epoch": 3.7642857142857142,
+      "grad_norm": 1.4275505542755127,
+      "learning_rate": 3.233333333333333e-05,
+      "loss": 0.3091,
+      "step": 5270
+    },
+    {
+      "epoch": 3.7714285714285714,
+      "grad_norm": 1.5278127193450928,
+      "learning_rate": 3.22962962962963e-05,
+      "loss": 0.3768,
+      "step": 5280
+    },
+    {
+      "epoch": 3.7785714285714285,
+      "grad_norm": 2.870471239089966,
+      "learning_rate": 3.225925925925926e-05,
+      "loss": 0.4264,
+      "step": 5290
+    },
+    {
+      "epoch": 3.7857142857142856,
+      "grad_norm": 1.4797722101211548,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 0.3598,
+      "step": 5300
+    },
+    {
+      "epoch": 3.7928571428571427,
+      "grad_norm": 1.6350576877593994,
+      "learning_rate": 3.2185185185185184e-05,
+      "loss": 0.2125,
+      "step": 5310
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.8790502548217773,
+      "learning_rate": 3.214814814814815e-05,
+      "loss": 0.2698,
+      "step": 5320
+    },
+    {
+      "epoch": 3.807142857142857,
+      "grad_norm": 1.3930083513259888,
+      "learning_rate": 3.2111111111111114e-05,
+      "loss": 0.3867,
+      "step": 5330
+    },
+    {
+      "epoch": 3.814285714285714,
+      "grad_norm": 1.7605199813842773,
+      "learning_rate": 3.2074074074074075e-05,
+      "loss": 0.3594,
+      "step": 5340
+    },
+    {
+      "epoch": 3.821428571428571,
+      "grad_norm": 2.3873794078826904,
+      "learning_rate": 3.203703703703704e-05,
+      "loss": 0.372,
+      "step": 5350
+    },
+    {
+      "epoch": 3.8285714285714287,
+      "grad_norm": 3.087186098098755,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.3964,
+      "step": 5360
+    },
+    {
+      "epoch": 3.835714285714286,
+      "grad_norm": 1.6758490800857544,
+      "learning_rate": 3.196296296296297e-05,
+      "loss": 0.3274,
+      "step": 5370
+    },
+    {
+      "epoch": 3.842857142857143,
+      "grad_norm": 1.184205412864685,
+      "learning_rate": 3.192592592592593e-05,
+      "loss": 0.277,
+      "step": 5380
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 2.1282460689544678,
+      "learning_rate": 3.188888888888889e-05,
+      "loss": 0.3283,
+      "step": 5390
+    },
+    {
+      "epoch": 3.857142857142857,
+      "grad_norm": 1.9244283437728882,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.2732,
+      "step": 5400
+    },
+    {
+      "epoch": 3.8642857142857143,
+      "grad_norm": 1.2328709363937378,
+      "learning_rate": 3.181481481481481e-05,
+      "loss": 0.2968,
+      "step": 5410
+    },
+    {
+      "epoch": 3.8714285714285714,
+      "grad_norm": 2.5490071773529053,
+      "learning_rate": 3.177777777777778e-05,
+      "loss": 0.3258,
+      "step": 5420
+    },
+    {
+      "epoch": 3.8785714285714286,
+      "grad_norm": 1.7774560451507568,
+      "learning_rate": 3.174074074074074e-05,
+      "loss": 0.3274,
+      "step": 5430
+    },
+    {
+      "epoch": 3.8857142857142857,
+      "grad_norm": 0.9900962710380554,
+      "learning_rate": 3.1703703703703705e-05,
+      "loss": 0.3361,
+      "step": 5440
+    },
+    {
+      "epoch": 3.892857142857143,
+      "grad_norm": 1.2809844017028809,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 0.3684,
+      "step": 5450
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.2611334323883057,
+      "learning_rate": 3.1629629629629634e-05,
+      "loss": 0.326,
+      "step": 5460
+    },
+    {
+      "epoch": 3.907142857142857,
+      "grad_norm": 2.49057936668396,
+      "learning_rate": 3.1592592592592596e-05,
+      "loss": 0.412,
+      "step": 5470
+    },
+    {
+      "epoch": 3.914285714285714,
+      "grad_norm": 1.6978118419647217,
+      "learning_rate": 3.155555555555556e-05,
+      "loss": 0.2177,
+      "step": 5480
+    },
+    {
+      "epoch": 3.9214285714285713,
+      "grad_norm": 1.847128987312317,
+      "learning_rate": 3.151851851851852e-05,
+      "loss": 0.3419,
+      "step": 5490
+    },
+    {
+      "epoch": 3.928571428571429,
+      "grad_norm": 1.6806657314300537,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.1479,
+      "step": 5500
+    },
+    {
+      "epoch": 3.935714285714286,
+      "grad_norm": 2.144227981567383,
+      "learning_rate": 3.144444444444445e-05,
+      "loss": 0.3098,
+      "step": 5510
+    },
+    {
+      "epoch": 3.942857142857143,
+      "grad_norm": 1.2945857048034668,
+      "learning_rate": 3.140740740740741e-05,
+      "loss": 0.269,
+      "step": 5520
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.8362900018692017,
+      "learning_rate": 3.137037037037037e-05,
+      "loss": 0.3065,
+      "step": 5530
+    },
+    {
+      "epoch": 3.9571428571428573,
+      "grad_norm": 1.9124987125396729,
+      "learning_rate": 3.1333333333333334e-05,
+      "loss": 0.2593,
+      "step": 5540
+    },
+    {
+      "epoch": 3.9642857142857144,
+      "grad_norm": 1.726523995399475,
+      "learning_rate": 3.1296296296296295e-05,
+      "loss": 0.3112,
+      "step": 5550
+    },
+    {
+      "epoch": 3.9714285714285715,
+      "grad_norm": 1.5914565324783325,
+      "learning_rate": 3.1259259259259264e-05,
+      "loss": 0.263,
+      "step": 5560
+    },
+    {
+      "epoch": 3.9785714285714286,
+      "grad_norm": 1.3533891439437866,
+      "learning_rate": 3.1222222222222225e-05,
+      "loss": 0.3852,
+      "step": 5570
+    },
+    {
+      "epoch": 3.9857142857142858,
+      "grad_norm": 2.1844253540039062,
+      "learning_rate": 3.118518518518519e-05,
+      "loss": 0.3761,
+      "step": 5580
+    },
+    {
+      "epoch": 3.992857142857143,
+      "grad_norm": 2.494920492172241,
+      "learning_rate": 3.114814814814815e-05,
+      "loss": 0.3882,
+      "step": 5590
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9914864897727966,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.3518,
+      "step": 5600
+    },
+    {
+      "epoch": 4.007142857142857,
+      "grad_norm": 1.6416865587234497,
+      "learning_rate": 3.107407407407408e-05,
+      "loss": 0.2688,
+      "step": 5610
+    },
+    {
+      "epoch": 4.014285714285714,
+      "grad_norm": 1.934449315071106,
+      "learning_rate": 3.103703703703704e-05,
+      "loss": 0.2385,
+      "step": 5620
+    },
+    {
+      "epoch": 4.021428571428571,
+      "grad_norm": 1.7663776874542236,
+      "learning_rate": 3.1e-05,
+      "loss": 0.3147,
+      "step": 5630
+    },
+    {
+      "epoch": 4.0285714285714285,
+      "grad_norm": 1.8457096815109253,
+      "learning_rate": 3.096296296296296e-05,
+      "loss": 0.2922,
+      "step": 5640
+    },
+    {
+      "epoch": 4.035714285714286,
+      "grad_norm": 1.133711338043213,
+      "learning_rate": 3.0925925925925924e-05,
+      "loss": 0.2291,
+      "step": 5650
+    },
+    {
+      "epoch": 4.042857142857143,
+      "grad_norm": 1.794723629951477,
+      "learning_rate": 3.088888888888889e-05,
+      "loss": 0.3204,
+      "step": 5660
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.966180443763733,
+      "learning_rate": 3.0851851851851854e-05,
+      "loss": 0.2757,
+      "step": 5670
+    },
+    {
+      "epoch": 4.057142857142857,
+      "grad_norm": 0.789313018321991,
+      "learning_rate": 3.0814814814814816e-05,
+      "loss": 0.3106,
+      "step": 5680
+    },
+    {
+      "epoch": 4.064285714285714,
+      "grad_norm": 1.4390606880187988,
+      "learning_rate": 3.077777777777778e-05,
+      "loss": 0.192,
+      "step": 5690
+    },
+    {
+      "epoch": 4.071428571428571,
+      "grad_norm": 1.8229310512542725,
+      "learning_rate": 3.074074074074074e-05,
+      "loss": 0.3802,
+      "step": 5700
+    },
+    {
+      "epoch": 4.078571428571428,
+      "grad_norm": 1.3065968751907349,
+      "learning_rate": 3.070370370370371e-05,
+      "loss": 0.2891,
+      "step": 5710
+    },
+    {
+      "epoch": 4.085714285714285,
+      "grad_norm": 1.5169206857681274,
+      "learning_rate": 3.066666666666667e-05,
+      "loss": 0.2818,
+      "step": 5720
+    },
+    {
+      "epoch": 4.0928571428571425,
+      "grad_norm": 1.8811321258544922,
+      "learning_rate": 3.062962962962963e-05,
+      "loss": 0.1845,
+      "step": 5730
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 2.2235770225524902,
+      "learning_rate": 3.059259259259259e-05,
+      "loss": 0.3671,
+      "step": 5740
+    },
+    {
+      "epoch": 4.107142857142857,
+      "grad_norm": 1.5675430297851562,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.3588,
+      "step": 5750
+    },
+    {
+      "epoch": 4.114285714285714,
+      "grad_norm": 1.3254741430282593,
+      "learning_rate": 3.0518518518518515e-05,
+      "loss": 0.3641,
+      "step": 5760
+    },
+    {
+      "epoch": 4.121428571428571,
+      "grad_norm": 2.601593017578125,
+      "learning_rate": 3.0481481481481484e-05,
+      "loss": 0.2704,
+      "step": 5770
+    },
+    {
+      "epoch": 4.128571428571428,
+      "grad_norm": 2.3631677627563477,
+      "learning_rate": 3.044444444444445e-05,
+      "loss": 0.2528,
+      "step": 5780
+    },
+    {
+      "epoch": 4.135714285714286,
+      "grad_norm": 1.4800968170166016,
+      "learning_rate": 3.0407407407407407e-05,
+      "loss": 0.263,
+      "step": 5790
+    },
+    {
+      "epoch": 4.142857142857143,
+      "grad_norm": 1.6989574432373047,
+      "learning_rate": 3.037037037037037e-05,
+      "loss": 0.2465,
+      "step": 5800
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.595765471458435,
+      "learning_rate": 3.0333333333333337e-05,
+      "loss": 0.3223,
+      "step": 5810
+    },
+    {
+      "epoch": 4.1571428571428575,
+      "grad_norm": 1.8895677328109741,
+      "learning_rate": 3.02962962962963e-05,
+      "loss": 0.3181,
+      "step": 5820
+    },
+    {
+      "epoch": 4.164285714285715,
+      "grad_norm": 1.147406816482544,
+      "learning_rate": 3.025925925925926e-05,
+      "loss": 0.2275,
+      "step": 5830
+    },
+    {
+      "epoch": 4.171428571428572,
+      "grad_norm": 3.310147523880005,
+      "learning_rate": 3.0222222222222225e-05,
+      "loss": 0.3615,
+      "step": 5840
+    },
+    {
+      "epoch": 4.178571428571429,
+      "grad_norm": 1.6138179302215576,
+      "learning_rate": 3.018518518518519e-05,
+      "loss": 0.3492,
+      "step": 5850
+    },
+    {
+      "epoch": 4.185714285714286,
+      "grad_norm": 1.9912358522415161,
+      "learning_rate": 3.0148148148148148e-05,
+      "loss": 0.3358,
+      "step": 5860
+    },
+    {
+      "epoch": 4.192857142857143,
+      "grad_norm": 2.2521820068359375,
+      "learning_rate": 3.0111111111111113e-05,
+      "loss": 0.2773,
+      "step": 5870
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.804829478263855,
+      "learning_rate": 3.0074074074074078e-05,
+      "loss": 0.3052,
+      "step": 5880
+    },
+    {
+      "epoch": 4.207142857142857,
+      "grad_norm": 1.0897246599197388,
+      "learning_rate": 3.0037037037037036e-05,
+      "loss": 0.3822,
+      "step": 5890
+    },
+    {
+      "epoch": 4.214285714285714,
+      "grad_norm": 1.337428331375122,
+      "learning_rate": 3e-05,
+      "loss": 0.3091,
+      "step": 5900
+    },
+    {
+      "epoch": 4.2214285714285715,
+      "grad_norm": 1.1409244537353516,
+      "learning_rate": 2.9962962962962966e-05,
+      "loss": 0.2002,
+      "step": 5910
+    },
+    {
+      "epoch": 4.228571428571429,
+      "grad_norm": 0.9190034866333008,
+      "learning_rate": 2.992592592592593e-05,
+      "loss": 0.3029,
+      "step": 5920
+    },
+    {
+      "epoch": 4.235714285714286,
+      "grad_norm": 1.7410012483596802,
+      "learning_rate": 2.988888888888889e-05,
+      "loss": 0.2361,
+      "step": 5930
+    },
+    {
+      "epoch": 4.242857142857143,
+      "grad_norm": 2.308295965194702,
+      "learning_rate": 2.9851851851851854e-05,
+      "loss": 0.3654,
+      "step": 5940
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.299177646636963,
+      "learning_rate": 2.981481481481482e-05,
+      "loss": 0.2346,
+      "step": 5950
+    },
+    {
+      "epoch": 4.257142857142857,
+      "grad_norm": 1.0352667570114136,
+      "learning_rate": 2.9777777777777777e-05,
+      "loss": 0.2331,
+      "step": 5960
+    },
+    {
+      "epoch": 4.264285714285714,
+      "grad_norm": 1.0682189464569092,
+      "learning_rate": 2.9740740740740742e-05,
+      "loss": 0.2456,
+      "step": 5970
+    },
+    {
+      "epoch": 4.271428571428571,
+      "grad_norm": 1.536718487739563,
+      "learning_rate": 2.9703703703703707e-05,
+      "loss": 0.1908,
+      "step": 5980
+    },
+    {
+      "epoch": 4.2785714285714285,
+      "grad_norm": 2.0448334217071533,
+      "learning_rate": 2.9666666666666672e-05,
+      "loss": 0.3399,
+      "step": 5990
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 2.205901622772217,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.1913,
+      "step": 6000
+    },
+    {
+      "epoch": 4.285714285714286,
+      "eval_loss": 0.36299219727516174,
+      "eval_rouge1": 0.9025,
+      "eval_rouge2": 0.8402,
+      "eval_rougeL": 0.8994,
+      "eval_runtime": 122.2765,
+      "eval_samples_per_second": 11.449,
+      "eval_steps_per_second": 5.725,
+      "step": 6000
+    },
+    {
+      "epoch": 4.292857142857143,
+      "grad_norm": 1.455069661140442,
+      "learning_rate": 2.9592592592592595e-05,
+      "loss": 0.2236,
+      "step": 6010
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.6218276023864746,
+      "learning_rate": 2.955555555555556e-05,
+      "loss": 0.2166,
+      "step": 6020
+    },
+    {
+      "epoch": 4.307142857142857,
+      "grad_norm": 1.4643278121948242,
+      "learning_rate": 2.9518518518518518e-05,
+      "loss": 0.2543,
+      "step": 6030
+    },
+    {
+      "epoch": 4.314285714285714,
+      "grad_norm": 1.9875061511993408,
+      "learning_rate": 2.9481481481481483e-05,
+      "loss": 0.275,
+      "step": 6040
+    },
+    {
+      "epoch": 4.321428571428571,
+      "grad_norm": 2.003077268600464,
+      "learning_rate": 2.9444444444444448e-05,
+      "loss": 0.3431,
+      "step": 6050
+    },
+    {
+      "epoch": 4.328571428571428,
+      "grad_norm": 1.332705020904541,
+      "learning_rate": 2.9407407407407413e-05,
+      "loss": 0.2546,
+      "step": 6060
+    },
+    {
+      "epoch": 4.335714285714285,
+      "grad_norm": 1.9161280393600464,
+      "learning_rate": 2.937037037037037e-05,
+      "loss": 0.2909,
+      "step": 6070
+    },
+    {
+      "epoch": 4.3428571428571425,
+      "grad_norm": 1.509238839149475,
+      "learning_rate": 2.9333333333333336e-05,
+      "loss": 0.253,
+      "step": 6080
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 2.238847255706787,
+      "learning_rate": 2.92962962962963e-05,
+      "loss": 0.2717,
+      "step": 6090
+    },
+    {
+      "epoch": 4.357142857142857,
+      "grad_norm": 1.9578133821487427,
+      "learning_rate": 2.925925925925926e-05,
+      "loss": 0.3407,
+      "step": 6100
+    },
+    {
+      "epoch": 4.364285714285714,
+      "grad_norm": 1.805828332901001,
+      "learning_rate": 2.9222222222222224e-05,
+      "loss": 0.1811,
+      "step": 6110
+    },
+    {
+      "epoch": 4.371428571428572,
+      "grad_norm": 2.9014134407043457,
+      "learning_rate": 2.918518518518519e-05,
+      "loss": 0.3934,
+      "step": 6120
+    },
+    {
+      "epoch": 4.378571428571428,
+      "grad_norm": 1.9857615232467651,
+      "learning_rate": 2.914814814814815e-05,
+      "loss": 0.2026,
+      "step": 6130
+    },
+    {
+      "epoch": 4.385714285714286,
+      "grad_norm": 2.3884503841400146,
+      "learning_rate": 2.9111111111111112e-05,
+      "loss": 0.2787,
+      "step": 6140
+    },
+    {
+      "epoch": 4.392857142857143,
+      "grad_norm": 2.298215866088867,
+      "learning_rate": 2.9074074074074077e-05,
+      "loss": 0.2765,
+      "step": 6150
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.1733076572418213,
+      "learning_rate": 2.9037037037037042e-05,
+      "loss": 0.3975,
+      "step": 6160
+    },
+    {
+      "epoch": 4.4071428571428575,
+      "grad_norm": 3.3003320693969727,
+      "learning_rate": 2.9e-05,
+      "loss": 0.4152,
+      "step": 6170
+    },
+    {
+      "epoch": 4.414285714285715,
+      "grad_norm": 1.5066970586776733,
+      "learning_rate": 2.8962962962962965e-05,
+      "loss": 0.345,
+      "step": 6180
+    },
+    {
+      "epoch": 4.421428571428572,
+      "grad_norm": 2.134096145629883,
+      "learning_rate": 2.892592592592593e-05,
+      "loss": 0.3154,
+      "step": 6190
+    },
+    {
+      "epoch": 4.428571428571429,
+      "grad_norm": 1.8306220769882202,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.2908,
+      "step": 6200
+    },
+    {
+      "epoch": 4.435714285714286,
+      "grad_norm": 1.4300037622451782,
+      "learning_rate": 2.8851851851851853e-05,
+      "loss": 0.342,
+      "step": 6210
+    },
+    {
+      "epoch": 4.442857142857143,
+      "grad_norm": 1.6552793979644775,
+      "learning_rate": 2.8814814814814818e-05,
+      "loss": 0.2856,
+      "step": 6220
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 2.188889265060425,
+      "learning_rate": 2.877777777777778e-05,
+      "loss": 0.25,
+      "step": 6230
+    },
+    {
+      "epoch": 4.457142857142857,
+      "grad_norm": 1.3003034591674805,
+      "learning_rate": 2.874074074074074e-05,
+      "loss": 0.2995,
+      "step": 6240
+    },
+    {
+      "epoch": 4.464285714285714,
+      "grad_norm": 1.834549903869629,
+      "learning_rate": 2.8703703703703706e-05,
+      "loss": 0.3726,
+      "step": 6250
+    },
+    {
+      "epoch": 4.4714285714285715,
+      "grad_norm": 1.9426199197769165,
+      "learning_rate": 2.8666666666666668e-05,
+      "loss": 0.2142,
+      "step": 6260
+    },
+    {
+      "epoch": 4.478571428571429,
+      "grad_norm": 1.5088646411895752,
+      "learning_rate": 2.862962962962963e-05,
+      "loss": 0.3584,
+      "step": 6270
+    },
+    {
+      "epoch": 4.485714285714286,
+      "grad_norm": 1.9997400045394897,
+      "learning_rate": 2.8592592592592594e-05,
+      "loss": 0.2402,
+      "step": 6280
+    },
+    {
+      "epoch": 4.492857142857143,
+      "grad_norm": 1.3831549882888794,
+      "learning_rate": 2.855555555555556e-05,
+      "loss": 0.312,
+      "step": 6290
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 2.013425588607788,
+      "learning_rate": 2.851851851851852e-05,
+      "loss": 0.2728,
+      "step": 6300
+    },
+    {
+      "epoch": 4.507142857142857,
+      "grad_norm": 1.1200778484344482,
+      "learning_rate": 2.8481481481481482e-05,
+      "loss": 0.3909,
+      "step": 6310
+    },
+    {
+      "epoch": 4.514285714285714,
+      "grad_norm": 0.8029781579971313,
+      "learning_rate": 2.8444444444444447e-05,
+      "loss": 0.3491,
+      "step": 6320
+    },
+    {
+      "epoch": 4.521428571428571,
+      "grad_norm": 1.4999722242355347,
+      "learning_rate": 2.840740740740741e-05,
+      "loss": 0.2583,
+      "step": 6330
+    },
+    {
+      "epoch": 4.5285714285714285,
+      "grad_norm": 1.8954156637191772,
+      "learning_rate": 2.837037037037037e-05,
+      "loss": 0.3971,
+      "step": 6340
+    },
+    {
+      "epoch": 4.535714285714286,
+      "grad_norm": 1.5697578191757202,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 0.3222,
+      "step": 6350
+    },
+    {
+      "epoch": 4.542857142857143,
+      "grad_norm": 0.9937646389007568,
+      "learning_rate": 2.8296296296296297e-05,
+      "loss": 0.3673,
+      "step": 6360
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.935511589050293,
+      "learning_rate": 2.8259259259259262e-05,
+      "loss": 0.2385,
+      "step": 6370
+    },
+    {
+      "epoch": 4.557142857142857,
+      "grad_norm": 1.8132340908050537,
+      "learning_rate": 2.8222222222222223e-05,
+      "loss": 0.226,
+      "step": 6380
+    },
+    {
+      "epoch": 4.564285714285714,
+      "grad_norm": 0.8551497459411621,
+      "learning_rate": 2.8185185185185185e-05,
+      "loss": 0.3874,
+      "step": 6390
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 2.0115785598754883,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.2328,
+      "step": 6400
+    },
+    {
+      "epoch": 4.578571428571428,
+      "grad_norm": 1.0582072734832764,
+      "learning_rate": 2.811111111111111e-05,
+      "loss": 0.3523,
+      "step": 6410
+    },
+    {
+      "epoch": 4.585714285714285,
+      "grad_norm": 1.3484958410263062,
+      "learning_rate": 2.8074074074074076e-05,
+      "loss": 0.2867,
+      "step": 6420
+    },
+    {
+      "epoch": 4.5928571428571425,
+      "grad_norm": 1.4483561515808105,
+      "learning_rate": 2.8037037037037038e-05,
+      "loss": 0.2623,
+      "step": 6430
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 2.2348268032073975,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.3953,
+      "step": 6440
+    },
+    {
+      "epoch": 4.607142857142857,
+      "grad_norm": 2.654326915740967,
+      "learning_rate": 2.7962962962962965e-05,
+      "loss": 0.3516,
+      "step": 6450
+    },
+    {
+      "epoch": 4.614285714285714,
+      "grad_norm": 0.8564252257347107,
+      "learning_rate": 2.7925925925925926e-05,
+      "loss": 0.2497,
+      "step": 6460
+    },
+    {
+      "epoch": 4.621428571428572,
+      "grad_norm": 2.7823233604431152,
+      "learning_rate": 2.788888888888889e-05,
+      "loss": 0.3975,
+      "step": 6470
+    },
+    {
+      "epoch": 4.628571428571428,
+      "grad_norm": 1.0915263891220093,
+      "learning_rate": 2.7851851851851853e-05,
+      "loss": 0.2574,
+      "step": 6480
+    },
+    {
+      "epoch": 4.635714285714286,
+      "grad_norm": 1.0459774732589722,
+      "learning_rate": 2.7814814814814814e-05,
+      "loss": 0.3426,
+      "step": 6490
+    },
+    {
+      "epoch": 4.642857142857143,
+      "grad_norm": 3.1720130443573,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.3155,
+      "step": 6500
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.499185562133789,
+      "learning_rate": 2.774074074074074e-05,
+      "loss": 0.4515,
+      "step": 6510
+    },
+    {
+      "epoch": 4.6571428571428575,
+      "grad_norm": 2.4211909770965576,
+      "learning_rate": 2.7703703703703706e-05,
+      "loss": 0.2963,
+      "step": 6520
+    },
+    {
+      "epoch": 4.664285714285715,
+      "grad_norm": 2.167006492614746,
+      "learning_rate": 2.7666666666666667e-05,
+      "loss": 0.2625,
+      "step": 6530
+    },
+    {
+      "epoch": 4.671428571428572,
+      "grad_norm": 1.8955094814300537,
+      "learning_rate": 2.7629629629629632e-05,
+      "loss": 0.3374,
+      "step": 6540
+    },
+    {
+      "epoch": 4.678571428571429,
+      "grad_norm": 0.9967934489250183,
+      "learning_rate": 2.7592592592592594e-05,
+      "loss": 0.1611,
+      "step": 6550
+    },
+    {
+      "epoch": 4.685714285714286,
+      "grad_norm": 1.007778525352478,
+      "learning_rate": 2.7555555555555555e-05,
+      "loss": 0.2516,
+      "step": 6560
+    },
+    {
+      "epoch": 4.692857142857143,
+      "grad_norm": 2.9705958366394043,
+      "learning_rate": 2.751851851851852e-05,
+      "loss": 0.3893,
+      "step": 6570
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 2.689723491668701,
+      "learning_rate": 2.7481481481481482e-05,
+      "loss": 0.2404,
+      "step": 6580
+    },
+    {
+      "epoch": 4.707142857142857,
+      "grad_norm": 2.095930337905884,
+      "learning_rate": 2.7444444444444443e-05,
+      "loss": 0.3239,
+      "step": 6590
+    },
+    {
+      "epoch": 4.714285714285714,
+      "grad_norm": 1.9235697984695435,
+      "learning_rate": 2.7407407407407408e-05,
+      "loss": 0.2779,
+      "step": 6600
+    },
+    {
+      "epoch": 4.7214285714285715,
+      "grad_norm": 3.329378843307495,
+      "learning_rate": 2.7370370370370373e-05,
+      "loss": 0.2791,
+      "step": 6610
+    },
+    {
+      "epoch": 4.728571428571429,
+      "grad_norm": 1.9044978618621826,
+      "learning_rate": 2.733333333333333e-05,
+      "loss": 0.3757,
+      "step": 6620
+    },
+    {
+      "epoch": 4.735714285714286,
+      "grad_norm": 2.207752227783203,
+      "learning_rate": 2.7296296296296296e-05,
+      "loss": 0.3391,
+      "step": 6630
+    },
+    {
+      "epoch": 4.742857142857143,
+      "grad_norm": 2.0488827228546143,
+      "learning_rate": 2.725925925925926e-05,
+      "loss": 0.396,
+      "step": 6640
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 2.425340414047241,
+      "learning_rate": 2.7222222222222223e-05,
+      "loss": 0.2871,
+      "step": 6650
+    },
+    {
+      "epoch": 4.757142857142857,
+      "grad_norm": 1.9408286809921265,
+      "learning_rate": 2.7185185185185184e-05,
+      "loss": 0.3144,
+      "step": 6660
+    },
+    {
+      "epoch": 4.764285714285714,
+      "grad_norm": 1.864397406578064,
+      "learning_rate": 2.714814814814815e-05,
+      "loss": 0.2685,
+      "step": 6670
+    },
+    {
+      "epoch": 4.771428571428571,
+      "grad_norm": 1.1838607788085938,
+      "learning_rate": 2.7111111111111114e-05,
+      "loss": 0.2751,
+      "step": 6680
+    },
+    {
+      "epoch": 4.7785714285714285,
+      "grad_norm": 2.26408052444458,
+      "learning_rate": 2.7074074074074072e-05,
+      "loss": 0.3158,
+      "step": 6690
+    },
+    {
+      "epoch": 4.785714285714286,
+      "grad_norm": 2.007145404815674,
+      "learning_rate": 2.7037037037037037e-05,
+      "loss": 0.1969,
+      "step": 6700
+    },
+    {
+      "epoch": 4.792857142857143,
+      "grad_norm": 2.5209295749664307,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.3022,
+      "step": 6710
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 2.3263044357299805,
+      "learning_rate": 2.696296296296296e-05,
+      "loss": 0.3799,
+      "step": 6720
+    },
+    {
+      "epoch": 4.807142857142857,
+      "grad_norm": 1.3880634307861328,
+      "learning_rate": 2.6925925925925925e-05,
+      "loss": 0.2829,
+      "step": 6730
+    },
+    {
+      "epoch": 4.814285714285714,
+      "grad_norm": 2.0264179706573486,
+      "learning_rate": 2.688888888888889e-05,
+      "loss": 0.2754,
+      "step": 6740
+    },
+    {
+      "epoch": 4.821428571428571,
+      "grad_norm": 1.6165140867233276,
+      "learning_rate": 2.6851851851851855e-05,
+      "loss": 0.3171,
+      "step": 6750
+    },
+    {
+      "epoch": 4.828571428571428,
+      "grad_norm": 1.6405526399612427,
+      "learning_rate": 2.6814814814814814e-05,
+      "loss": 0.4082,
+      "step": 6760
+    },
+    {
+      "epoch": 4.835714285714285,
+      "grad_norm": 1.6864060163497925,
+      "learning_rate": 2.677777777777778e-05,
+      "loss": 0.2026,
+      "step": 6770
+    },
+    {
+      "epoch": 4.8428571428571425,
+      "grad_norm": 1.4906965494155884,
+      "learning_rate": 2.6740740740740743e-05,
+      "loss": 0.2582,
+      "step": 6780
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.2227530479431152,
+      "learning_rate": 2.67037037037037e-05,
+      "loss": 0.185,
+      "step": 6790
+    },
+    {
+      "epoch": 4.857142857142857,
+      "grad_norm": 1.2606697082519531,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.2651,
+      "step": 6800
+    },
+    {
+      "epoch": 4.864285714285714,
+      "grad_norm": 2.3722660541534424,
+      "learning_rate": 2.662962962962963e-05,
+      "loss": 0.2746,
+      "step": 6810
+    },
+    {
+      "epoch": 4.871428571428572,
+      "grad_norm": 1.8622608184814453,
+      "learning_rate": 2.659259259259259e-05,
+      "loss": 0.3473,
+      "step": 6820
+    },
+    {
+      "epoch": 4.878571428571428,
+      "grad_norm": 1.3814878463745117,
+      "learning_rate": 2.6555555555555555e-05,
+      "loss": 0.2706,
+      "step": 6830
+    },
+    {
+      "epoch": 4.885714285714286,
+      "grad_norm": 2.013650894165039,
+      "learning_rate": 2.651851851851852e-05,
+      "loss": 0.2802,
+      "step": 6840
+    },
+    {
+      "epoch": 4.892857142857143,
+      "grad_norm": 1.467282772064209,
+      "learning_rate": 2.6481481481481485e-05,
+      "loss": 0.3158,
+      "step": 6850
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.3019797801971436,
+      "learning_rate": 2.6444444444444443e-05,
+      "loss": 0.2012,
+      "step": 6860
+    },
+    {
+      "epoch": 4.9071428571428575,
+      "grad_norm": 1.1120600700378418,
+      "learning_rate": 2.6407407407407408e-05,
+      "loss": 0.1385,
+      "step": 6870
+    },
+    {
+      "epoch": 4.914285714285715,
+      "grad_norm": 1.470406413078308,
+      "learning_rate": 2.6370370370370373e-05,
+      "loss": 0.3014,
+      "step": 6880
+    },
+    {
+      "epoch": 4.921428571428572,
+      "grad_norm": 2.237767457962036,
+      "learning_rate": 2.633333333333333e-05,
+      "loss": 0.2677,
+      "step": 6890
+    },
+    {
+      "epoch": 4.928571428571429,
+      "grad_norm": 1.3994693756103516,
+      "learning_rate": 2.6296296296296296e-05,
+      "loss": 0.4261,
+      "step": 6900
+    },
+    {
+      "epoch": 4.935714285714286,
+      "grad_norm": 2.21905517578125,
+      "learning_rate": 2.625925925925926e-05,
+      "loss": 0.3701,
+      "step": 6910
+    },
+    {
+      "epoch": 4.942857142857143,
+      "grad_norm": 2.8682186603546143,
+      "learning_rate": 2.6222222222222226e-05,
+      "loss": 0.4047,
+      "step": 6920
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.9691041707992554,
+      "learning_rate": 2.6185185185185184e-05,
+      "loss": 0.2735,
+      "step": 6930
+    },
+    {
+      "epoch": 4.957142857142857,
+      "grad_norm": 1.7553354501724243,
+      "learning_rate": 2.614814814814815e-05,
+      "loss": 0.2381,
+      "step": 6940
+    },
+    {
+      "epoch": 4.964285714285714,
+      "grad_norm": 1.7930738925933838,
+      "learning_rate": 2.6111111111111114e-05,
+      "loss": 0.2838,
+      "step": 6950
+    },
+    {
+      "epoch": 4.9714285714285715,
+      "grad_norm": 2.4153687953948975,
+      "learning_rate": 2.6074074074074072e-05,
+      "loss": 0.4002,
+      "step": 6960
+    },
+    {
+      "epoch": 4.978571428571429,
+      "grad_norm": 1.392898678779602,
+      "learning_rate": 2.6037037037037037e-05,
+      "loss": 0.248,
+      "step": 6970
+    },
+    {
+      "epoch": 4.985714285714286,
+      "grad_norm": 1.7113401889801025,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.2871,
+      "step": 6980
+    },
+    {
+      "epoch": 4.992857142857143,
+      "grad_norm": 2.4877359867095947,
+      "learning_rate": 2.5962962962962967e-05,
+      "loss": 0.2443,
+      "step": 6990
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.7225149869918823,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.186,
+      "step": 7000
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.3584790527820587,
+      "eval_rouge1": 0.9047,
+      "eval_rouge2": 0.8434,
+      "eval_rougeL": 0.9018,
+      "eval_runtime": 122.2903,
+      "eval_samples_per_second": 11.448,
+      "eval_steps_per_second": 5.724,
+      "step": 7000
+    },
+    {
+      "epoch": 5.007142857142857,
+      "grad_norm": 2.1430020332336426,
+      "learning_rate": 2.588888888888889e-05,
+      "loss": 0.3477,
+      "step": 7010
+    },
+    {
+      "epoch": 5.014285714285714,
+      "grad_norm": 0.958677351474762,
+      "learning_rate": 2.5851851851851855e-05,
+      "loss": 0.2474,
+      "step": 7020
+    },
+    {
+      "epoch": 5.021428571428571,
+      "grad_norm": 2.315269947052002,
+      "learning_rate": 2.5814814814814813e-05,
+      "loss": 0.2786,
+      "step": 7030
+    },
+    {
+      "epoch": 5.0285714285714285,
+      "grad_norm": 1.3595519065856934,
+      "learning_rate": 2.5777777777777778e-05,
+      "loss": 0.2286,
+      "step": 7040
+    },
+    {
+      "epoch": 5.035714285714286,
+      "grad_norm": 1.44675874710083,
+      "learning_rate": 2.5740740740740743e-05,
+      "loss": 0.2679,
+      "step": 7050
+    },
+    {
+      "epoch": 5.042857142857143,
+      "grad_norm": 1.754285454750061,
+      "learning_rate": 2.5703703703703708e-05,
+      "loss": 0.196,
+      "step": 7060
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 2.9333369731903076,
+      "learning_rate": 2.5666666666666666e-05,
+      "loss": 0.1694,
+      "step": 7070
+    },
+    {
+      "epoch": 5.057142857142857,
+      "grad_norm": 2.6653859615325928,
+      "learning_rate": 2.562962962962963e-05,
+      "loss": 0.2642,
+      "step": 7080
+    },
+    {
+      "epoch": 5.064285714285714,
+      "grad_norm": 1.8362854719161987,
+      "learning_rate": 2.5592592592592596e-05,
+      "loss": 0.3614,
+      "step": 7090
+    },
+    {
+      "epoch": 5.071428571428571,
+      "grad_norm": 1.427701473236084,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 0.2351,
+      "step": 7100
+    },
+    {
+      "epoch": 5.078571428571428,
+      "grad_norm": 2.3684027194976807,
+      "learning_rate": 2.551851851851852e-05,
+      "loss": 0.2803,
+      "step": 7110
+    },
+    {
+      "epoch": 5.085714285714285,
+      "grad_norm": 1.5823931694030762,
+      "learning_rate": 2.5481481481481484e-05,
+      "loss": 0.2749,
+      "step": 7120
+    },
+    {
+      "epoch": 5.0928571428571425,
+      "grad_norm": 1.6682019233703613,
+      "learning_rate": 2.5444444444444442e-05,
+      "loss": 0.3219,
+      "step": 7130
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 1.7803760766983032,
+      "learning_rate": 2.5407407407407407e-05,
+      "loss": 0.2553,
+      "step": 7140
+    },
+    {
+      "epoch": 5.107142857142857,
+      "grad_norm": 1.945063591003418,
+      "learning_rate": 2.5370370370370372e-05,
+      "loss": 0.1739,
+      "step": 7150
+    },
+    {
+      "epoch": 5.114285714285714,
+      "grad_norm": 1.308371663093567,
+      "learning_rate": 2.5333333333333337e-05,
+      "loss": 0.2605,
+      "step": 7160
+    },
+    {
+      "epoch": 5.121428571428571,
+      "grad_norm": 1.906160593032837,
+      "learning_rate": 2.5296296296296295e-05,
+      "loss": 0.2071,
+      "step": 7170
+    },
+    {
+      "epoch": 5.128571428571428,
+      "grad_norm": 1.6239346265792847,
+      "learning_rate": 2.525925925925926e-05,
+      "loss": 0.2054,
+      "step": 7180
+    },
+    {
+      "epoch": 5.135714285714286,
+      "grad_norm": 1.6175967454910278,
+      "learning_rate": 2.5222222222222225e-05,
+      "loss": 0.2266,
+      "step": 7190
+    },
+    {
+      "epoch": 5.142857142857143,
+      "grad_norm": 1.938736915588379,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 0.2932,
+      "step": 7200
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 1.7323144674301147,
+      "learning_rate": 2.5148148148148148e-05,
+      "loss": 0.2762,
+      "step": 7210
+    },
+    {
+      "epoch": 5.1571428571428575,
+      "grad_norm": 1.859667181968689,
+      "learning_rate": 2.5111111111111113e-05,
+      "loss": 0.3213,
+      "step": 7220
+    },
+    {
+      "epoch": 5.164285714285715,
+      "grad_norm": 1.22067391872406,
+      "learning_rate": 2.5074074074074078e-05,
+      "loss": 0.2246,
+      "step": 7230
+    },
+    {
+      "epoch": 5.171428571428572,
+      "grad_norm": 0.9384840726852417,
+      "learning_rate": 2.5037037037037036e-05,
+      "loss": 0.3364,
+      "step": 7240
+    },
+    {
+      "epoch": 5.178571428571429,
+      "grad_norm": 1.4494845867156982,
+      "learning_rate": 2.5e-05,
+      "loss": 0.332,
+      "step": 7250
+    },
+    {
+      "epoch": 5.185714285714286,
+      "grad_norm": 2.3436357975006104,
+      "learning_rate": 2.4962962962962963e-05,
+      "loss": 0.1456,
+      "step": 7260
+    },
+    {
+      "epoch": 5.192857142857143,
+      "grad_norm": 1.0446144342422485,
+      "learning_rate": 2.4925925925925928e-05,
+      "loss": 0.1995,
+      "step": 7270
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 2.325575113296509,
+      "learning_rate": 2.488888888888889e-05,
+      "loss": 0.3068,
+      "step": 7280
+    },
+    {
+      "epoch": 5.207142857142857,
+      "grad_norm": 2.100825309753418,
+      "learning_rate": 2.4851851851851854e-05,
+      "loss": 0.2659,
+      "step": 7290
+    },
+    {
+      "epoch": 5.214285714285714,
+      "grad_norm": 2.6580276489257812,
+      "learning_rate": 2.4814814814814816e-05,
+      "loss": 0.2872,
+      "step": 7300
+    },
+    {
+      "epoch": 5.2214285714285715,
+      "grad_norm": 2.505577564239502,
+      "learning_rate": 2.477777777777778e-05,
+      "loss": 0.2574,
+      "step": 7310
+    },
+    {
+      "epoch": 5.228571428571429,
+      "grad_norm": 1.4997559785842896,
+      "learning_rate": 2.4740740740740742e-05,
+      "loss": 0.2192,
+      "step": 7320
+    },
+    {
+      "epoch": 5.235714285714286,
+      "grad_norm": 1.9084120988845825,
+      "learning_rate": 2.4703703703703704e-05,
+      "loss": 0.2836,
+      "step": 7330
+    },
+    {
+      "epoch": 5.242857142857143,
+      "grad_norm": 1.1388484239578247,
+      "learning_rate": 2.466666666666667e-05,
+      "loss": 0.2426,
+      "step": 7340
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.0559568405151367,
+      "learning_rate": 2.462962962962963e-05,
+      "loss": 0.344,
+      "step": 7350
+    },
+    {
+      "epoch": 5.257142857142857,
+      "grad_norm": 1.4024419784545898,
+      "learning_rate": 2.4592592592592595e-05,
+      "loss": 0.2121,
+      "step": 7360
+    },
+    {
+      "epoch": 5.264285714285714,
+      "grad_norm": 1.4338841438293457,
+      "learning_rate": 2.4555555555555557e-05,
+      "loss": 0.3329,
+      "step": 7370
+    },
+    {
+      "epoch": 5.271428571428571,
+      "grad_norm": 1.4188106060028076,
+      "learning_rate": 2.451851851851852e-05,
+      "loss": 0.2479,
+      "step": 7380
+    },
+    {
+      "epoch": 5.2785714285714285,
+      "grad_norm": 1.4320842027664185,
+      "learning_rate": 2.4481481481481483e-05,
+      "loss": 0.156,
+      "step": 7390
+    },
+    {
+      "epoch": 5.285714285714286,
+      "grad_norm": 3.022641181945801,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.1962,
+      "step": 7400
+    },
+    {
+      "epoch": 5.292857142857143,
+      "grad_norm": 2.3267366886138916,
+      "learning_rate": 2.440740740740741e-05,
+      "loss": 0.2713,
+      "step": 7410
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 2.685345411300659,
+      "learning_rate": 2.437037037037037e-05,
+      "loss": 0.3345,
+      "step": 7420
+    },
+    {
+      "epoch": 5.307142857142857,
+      "grad_norm": 0.9320240020751953,
+      "learning_rate": 2.4333333333333336e-05,
+      "loss": 0.3758,
+      "step": 7430
+    },
+    {
+      "epoch": 5.314285714285714,
+      "grad_norm": 1.8067562580108643,
+      "learning_rate": 2.4296296296296298e-05,
+      "loss": 0.2958,
+      "step": 7440
+    },
+    {
+      "epoch": 5.321428571428571,
+      "grad_norm": 1.5514296293258667,
+      "learning_rate": 2.425925925925926e-05,
+      "loss": 0.3268,
+      "step": 7450
+    },
+    {
+      "epoch": 5.328571428571428,
+      "grad_norm": 1.684311032295227,
+      "learning_rate": 2.4222222222222224e-05,
+      "loss": 0.2947,
+      "step": 7460
+    },
+    {
+      "epoch": 5.335714285714285,
+      "grad_norm": 2.0809545516967773,
+      "learning_rate": 2.4185185185185186e-05,
+      "loss": 0.2928,
+      "step": 7470
+    },
+    {
+      "epoch": 5.3428571428571425,
+      "grad_norm": 2.5362987518310547,
+      "learning_rate": 2.414814814814815e-05,
+      "loss": 0.1962,
+      "step": 7480
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.636965274810791,
+      "learning_rate": 2.4111111111111113e-05,
+      "loss": 0.1694,
+      "step": 7490
+    },
+    {
+      "epoch": 5.357142857142857,
+      "grad_norm": 2.1662261486053467,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.3111,
+      "step": 7500
+    },
+    {
+      "epoch": 5.364285714285714,
+      "grad_norm": 1.749324083328247,
+      "learning_rate": 2.403703703703704e-05,
+      "loss": 0.2521,
+      "step": 7510
+    },
+    {
+      "epoch": 5.371428571428572,
+      "grad_norm": 2.3572323322296143,
+      "learning_rate": 2.4e-05,
+      "loss": 0.1527,
+      "step": 7520
+    },
+    {
+      "epoch": 5.378571428571428,
+      "grad_norm": 1.274588942527771,
+      "learning_rate": 2.3962962962962966e-05,
+      "loss": 0.2757,
+      "step": 7530
+    },
+    {
+      "epoch": 5.385714285714286,
+      "grad_norm": 1.2197136878967285,
+      "learning_rate": 2.3925925925925927e-05,
+      "loss": 0.2288,
+      "step": 7540
+    },
+    {
+      "epoch": 5.392857142857143,
+      "grad_norm": 1.6061832904815674,
+      "learning_rate": 2.3888888888888892e-05,
+      "loss": 0.3292,
+      "step": 7550
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.8271028995513916,
+      "learning_rate": 2.3851851851851854e-05,
+      "loss": 0.2392,
+      "step": 7560
+    },
+    {
+      "epoch": 5.4071428571428575,
+      "grad_norm": 1.8294018507003784,
+      "learning_rate": 2.3814814814814815e-05,
+      "loss": 0.2554,
+      "step": 7570
+    },
+    {
+      "epoch": 5.414285714285715,
+      "grad_norm": 1.253556728363037,
+      "learning_rate": 2.377777777777778e-05,
+      "loss": 0.2008,
+      "step": 7580
+    },
+    {
+      "epoch": 5.421428571428572,
+      "grad_norm": 1.1980758905410767,
+      "learning_rate": 2.3740740740740742e-05,
+      "loss": 0.265,
+      "step": 7590
+    },
+    {
+      "epoch": 5.428571428571429,
+      "grad_norm": 1.5337406396865845,
+      "learning_rate": 2.3703703703703707e-05,
+      "loss": 0.4126,
+      "step": 7600
+    },
+    {
+      "epoch": 5.435714285714286,
+      "grad_norm": 2.981381893157959,
+      "learning_rate": 2.3666666666666668e-05,
+      "loss": 0.3554,
+      "step": 7610
+    },
+    {
+      "epoch": 5.442857142857143,
+      "grad_norm": 1.927241325378418,
+      "learning_rate": 2.3629629629629633e-05,
+      "loss": 0.3148,
+      "step": 7620
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 1.0788408517837524,
+      "learning_rate": 2.3592592592592595e-05,
+      "loss": 0.2421,
+      "step": 7630
+    },
+    {
+      "epoch": 5.457142857142857,
+      "grad_norm": 1.250436782836914,
+      "learning_rate": 2.3555555555555556e-05,
+      "loss": 0.2797,
+      "step": 7640
+    },
+    {
+      "epoch": 5.464285714285714,
+      "grad_norm": 1.2195000648498535,
+      "learning_rate": 2.351851851851852e-05,
+      "loss": 0.1702,
+      "step": 7650
+    },
+    {
+      "epoch": 5.4714285714285715,
+      "grad_norm": 1.773098349571228,
+      "learning_rate": 2.3481481481481483e-05,
+      "loss": 0.2383,
+      "step": 7660
+    },
+    {
+      "epoch": 5.478571428571429,
+      "grad_norm": 1.540499210357666,
+      "learning_rate": 2.3444444444444448e-05,
+      "loss": 0.2741,
+      "step": 7670
+    },
+    {
+      "epoch": 5.485714285714286,
+      "grad_norm": 1.3515613079071045,
+      "learning_rate": 2.340740740740741e-05,
+      "loss": 0.4365,
+      "step": 7680
+    },
+    {
+      "epoch": 5.492857142857143,
+      "grad_norm": 1.5094635486602783,
+      "learning_rate": 2.337037037037037e-05,
+      "loss": 0.2777,
+      "step": 7690
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 1.123542070388794,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.3406,
+      "step": 7700
+    },
+    {
+      "epoch": 5.507142857142857,
+      "grad_norm": 1.0701942443847656,
+      "learning_rate": 2.3296296296296297e-05,
+      "loss": 0.2499,
+      "step": 7710
+    },
+    {
+      "epoch": 5.514285714285714,
+      "grad_norm": 1.270992636680603,
+      "learning_rate": 2.3259259259259262e-05,
+      "loss": 0.2044,
+      "step": 7720
+    },
+    {
+      "epoch": 5.521428571428571,
+      "grad_norm": 1.5586347579956055,
+      "learning_rate": 2.3222222222222224e-05,
+      "loss": 0.2573,
+      "step": 7730
+    },
+    {
+      "epoch": 5.5285714285714285,
+      "grad_norm": 0.9162809252738953,
+      "learning_rate": 2.318518518518519e-05,
+      "loss": 0.2245,
+      "step": 7740
+    },
+    {
+      "epoch": 5.535714285714286,
+      "grad_norm": 1.7767843008041382,
+      "learning_rate": 2.314814814814815e-05,
+      "loss": 0.276,
+      "step": 7750
+    },
+    {
+      "epoch": 5.542857142857143,
+      "grad_norm": 2.538541316986084,
+      "learning_rate": 2.3111111111111112e-05,
+      "loss": 0.3448,
+      "step": 7760
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 1.5738705396652222,
+      "learning_rate": 2.3074074074074077e-05,
+      "loss": 0.3023,
+      "step": 7770
+    },
+    {
+      "epoch": 5.557142857142857,
+      "grad_norm": 0.9919751286506653,
+      "learning_rate": 2.303703703703704e-05,
+      "loss": 0.2979,
+      "step": 7780
+    },
+    {
+      "epoch": 5.564285714285714,
+      "grad_norm": 1.079817771911621,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.3141,
+      "step": 7790
+    },
+    {
+      "epoch": 5.571428571428571,
+      "grad_norm": 1.709007978439331,
+      "learning_rate": 2.2962962962962965e-05,
+      "loss": 0.2452,
+      "step": 7800
+    },
+    {
+      "epoch": 5.578571428571428,
+      "grad_norm": 0.9877552390098572,
+      "learning_rate": 2.2925925925925927e-05,
+      "loss": 0.2796,
+      "step": 7810
+    },
+    {
+      "epoch": 5.585714285714285,
+      "grad_norm": 1.9676953554153442,
+      "learning_rate": 2.288888888888889e-05,
+      "loss": 0.2314,
+      "step": 7820
+    },
+    {
+      "epoch": 5.5928571428571425,
+      "grad_norm": 1.778275966644287,
+      "learning_rate": 2.2851851851851853e-05,
+      "loss": 0.3033,
+      "step": 7830
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.9746494889259338,
+      "learning_rate": 2.2814814814814818e-05,
+      "loss": 0.2459,
+      "step": 7840
+    },
+    {
+      "epoch": 5.607142857142857,
+      "grad_norm": 2.0238702297210693,
+      "learning_rate": 2.277777777777778e-05,
+      "loss": 0.2494,
+      "step": 7850
+    },
+    {
+      "epoch": 5.614285714285714,
+      "grad_norm": 1.2345530986785889,
+      "learning_rate": 2.2740740740740744e-05,
+      "loss": 0.4614,
+      "step": 7860
+    },
+    {
+      "epoch": 5.621428571428572,
+      "grad_norm": 0.9835256338119507,
+      "learning_rate": 2.2703703703703706e-05,
+      "loss": 0.3519,
+      "step": 7870
+    },
+    {
+      "epoch": 5.628571428571428,
+      "grad_norm": 1.9753897190093994,
+      "learning_rate": 2.2666666666666668e-05,
+      "loss": 0.2895,
+      "step": 7880
+    },
+    {
+      "epoch": 5.635714285714286,
+      "grad_norm": 1.7247217893600464,
+      "learning_rate": 2.2629629629629633e-05,
+      "loss": 0.1994,
+      "step": 7890
+    },
+    {
+      "epoch": 5.642857142857143,
+      "grad_norm": 1.8406201601028442,
+      "learning_rate": 2.2592592592592594e-05,
+      "loss": 0.1872,
+      "step": 7900
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 1.4785393476486206,
+      "learning_rate": 2.255555555555556e-05,
+      "loss": 0.2811,
+      "step": 7910
+    },
+    {
+      "epoch": 5.6571428571428575,
+      "grad_norm": 2.23652982711792,
+      "learning_rate": 2.251851851851852e-05,
+      "loss": 0.3071,
+      "step": 7920
+    },
+    {
+      "epoch": 5.664285714285715,
+      "grad_norm": 1.9096837043762207,
+      "learning_rate": 2.2481481481481486e-05,
+      "loss": 0.2115,
+      "step": 7930
+    },
+    {
+      "epoch": 5.671428571428572,
+      "grad_norm": 2.0808775424957275,
+      "learning_rate": 2.2444444444444447e-05,
+      "loss": 0.3923,
+      "step": 7940
+    },
+    {
+      "epoch": 5.678571428571429,
+      "grad_norm": 1.5935535430908203,
+      "learning_rate": 2.240740740740741e-05,
+      "loss": 0.3461,
+      "step": 7950
+    },
+    {
+      "epoch": 5.685714285714286,
+      "grad_norm": 1.1959024667739868,
+      "learning_rate": 2.2370370370370374e-05,
+      "loss": 0.2016,
+      "step": 7960
+    },
+    {
+      "epoch": 5.692857142857143,
+      "grad_norm": 1.0776904821395874,
+      "learning_rate": 2.2333333333333335e-05,
+      "loss": 0.3476,
+      "step": 7970
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 1.884531855583191,
+      "learning_rate": 2.2296296296296297e-05,
+      "loss": 0.2861,
+      "step": 7980
+    },
+    {
+      "epoch": 5.707142857142857,
+      "grad_norm": 1.2476330995559692,
+      "learning_rate": 2.2259259259259262e-05,
+      "loss": 0.2152,
+      "step": 7990
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 2.106348752975464,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.3022,
+      "step": 8000
+    },
+    {
+      "epoch": 5.714285714285714,
+      "eval_loss": 0.3492221236228943,
+      "eval_rouge1": 0.9062,
+      "eval_rouge2": 0.8456,
+      "eval_rougeL": 0.9033,
+      "eval_runtime": 122.1433,
+      "eval_samples_per_second": 11.462,
+      "eval_steps_per_second": 5.731,
+      "step": 8000
+    },
+    {
+      "epoch": 5.7214285714285715,
+      "grad_norm": 1.9454623460769653,
+      "learning_rate": 2.2185185185185188e-05,
+      "loss": 0.232,
+      "step": 8010
+    },
+    {
+      "epoch": 5.728571428571429,
+      "grad_norm": 1.9178905487060547,
+      "learning_rate": 2.214814814814815e-05,
+      "loss": 0.2278,
+      "step": 8020
+    },
+    {
+      "epoch": 5.735714285714286,
+      "grad_norm": 1.6279345750808716,
+      "learning_rate": 2.211111111111111e-05,
+      "loss": 0.2423,
+      "step": 8030
+    },
+    {
+      "epoch": 5.742857142857143,
+      "grad_norm": 2.7422447204589844,
+      "learning_rate": 2.2074074074074076e-05,
+      "loss": 0.3129,
+      "step": 8040
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.7606775760650635,
+      "learning_rate": 2.2037037037037038e-05,
+      "loss": 0.217,
+      "step": 8050
+    },
+    {
+      "epoch": 5.757142857142857,
+      "grad_norm": 2.970276355743408,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.3246,
+      "step": 8060
+    },
+    {
+      "epoch": 5.764285714285714,
+      "grad_norm": 1.6729111671447754,
+      "learning_rate": 2.1962962962962964e-05,
+      "loss": 0.224,
+      "step": 8070
+    },
+    {
+      "epoch": 5.771428571428571,
+      "grad_norm": 2.103708267211914,
+      "learning_rate": 2.1925925925925926e-05,
+      "loss": 0.2256,
+      "step": 8080
+    },
+    {
+      "epoch": 5.7785714285714285,
+      "grad_norm": 1.7059235572814941,
+      "learning_rate": 2.188888888888889e-05,
+      "loss": 0.2986,
+      "step": 8090
+    },
+    {
+      "epoch": 5.785714285714286,
+      "grad_norm": 1.6239415407180786,
+      "learning_rate": 2.1851851851851852e-05,
+      "loss": 0.3007,
+      "step": 8100
+    },
+    {
+      "epoch": 5.792857142857143,
+      "grad_norm": 1.5316799879074097,
+      "learning_rate": 2.1814814814814817e-05,
+      "loss": 0.2295,
+      "step": 8110
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.9283231496810913,
+      "learning_rate": 2.177777777777778e-05,
+      "loss": 0.1624,
+      "step": 8120
+    },
+    {
+      "epoch": 5.807142857142857,
+      "grad_norm": 1.4230540990829468,
+      "learning_rate": 2.174074074074074e-05,
+      "loss": 0.1686,
+      "step": 8130
+    },
+    {
+      "epoch": 5.814285714285714,
+      "grad_norm": 1.8694360256195068,
+      "learning_rate": 2.1703703703703705e-05,
+      "loss": 0.3416,
+      "step": 8140
+    },
+    {
+      "epoch": 5.821428571428571,
+      "grad_norm": 2.144221782684326,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 0.2471,
+      "step": 8150
+    },
+    {
+      "epoch": 5.828571428571428,
+      "grad_norm": 2.5672965049743652,
+      "learning_rate": 2.162962962962963e-05,
+      "loss": 0.3354,
+      "step": 8160
+    },
+    {
+      "epoch": 5.835714285714285,
+      "grad_norm": 1.093578577041626,
+      "learning_rate": 2.1592592592592594e-05,
+      "loss": 0.262,
+      "step": 8170
+    },
+    {
+      "epoch": 5.8428571428571425,
+      "grad_norm": 0.7076272368431091,
+      "learning_rate": 2.1555555555555555e-05,
+      "loss": 0.254,
+      "step": 8180
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 2.2301125526428223,
+      "learning_rate": 2.151851851851852e-05,
+      "loss": 0.1906,
+      "step": 8190
+    },
+    {
+      "epoch": 5.857142857142857,
+      "grad_norm": 1.704037070274353,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 0.2802,
+      "step": 8200
+    },
+    {
+      "epoch": 5.864285714285714,
+      "grad_norm": 1.4877769947052002,
+      "learning_rate": 2.1444444444444443e-05,
+      "loss": 0.3327,
+      "step": 8210
+    },
+    {
+      "epoch": 5.871428571428572,
+      "grad_norm": 1.436059594154358,
+      "learning_rate": 2.1407407407407408e-05,
+      "loss": 0.271,
+      "step": 8220
+    },
+    {
+      "epoch": 5.878571428571428,
+      "grad_norm": 1.357176661491394,
+      "learning_rate": 2.137037037037037e-05,
+      "loss": 0.2481,
+      "step": 8230
+    },
+    {
+      "epoch": 5.885714285714286,
+      "grad_norm": 1.846593976020813,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 0.2641,
+      "step": 8240
+    },
+    {
+      "epoch": 5.892857142857143,
+      "grad_norm": 2.4631927013397217,
+      "learning_rate": 2.1296296296296296e-05,
+      "loss": 0.2832,
+      "step": 8250
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 1.8715349435806274,
+      "learning_rate": 2.1259259259259258e-05,
+      "loss": 0.4157,
+      "step": 8260
+    },
+    {
+      "epoch": 5.9071428571428575,
+      "grad_norm": 2.3173437118530273,
+      "learning_rate": 2.1222222222222223e-05,
+      "loss": 0.353,
+      "step": 8270
+    },
+    {
+      "epoch": 5.914285714285715,
+      "grad_norm": 2.049422025680542,
+      "learning_rate": 2.1185185185185184e-05,
+      "loss": 0.2613,
+      "step": 8280
+    },
+    {
+      "epoch": 5.921428571428572,
+      "grad_norm": 1.281841516494751,
+      "learning_rate": 2.114814814814815e-05,
+      "loss": 0.2287,
+      "step": 8290
+    },
+    {
+      "epoch": 5.928571428571429,
+      "grad_norm": 1.007407546043396,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 0.2139,
+      "step": 8300
+    },
+    {
+      "epoch": 5.935714285714286,
+      "grad_norm": 1.8036701679229736,
+      "learning_rate": 2.1074074074074072e-05,
+      "loss": 0.2511,
+      "step": 8310
+    },
+    {
+      "epoch": 5.942857142857143,
+      "grad_norm": 0.9559861421585083,
+      "learning_rate": 2.1037037037037037e-05,
+      "loss": 0.3371,
+      "step": 8320
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 2.136070489883423,
+      "learning_rate": 2.1e-05,
+      "loss": 0.2321,
+      "step": 8330
+    },
+    {
+      "epoch": 5.957142857142857,
+      "grad_norm": 1.2442055940628052,
+      "learning_rate": 2.0962962962962964e-05,
+      "loss": 0.1819,
+      "step": 8340
+    },
+    {
+      "epoch": 5.964285714285714,
+      "grad_norm": 2.0479979515075684,
+      "learning_rate": 2.0925925925925925e-05,
+      "loss": 0.3796,
+      "step": 8350
+    },
+    {
+      "epoch": 5.9714285714285715,
+      "grad_norm": 1.6974670886993408,
+      "learning_rate": 2.088888888888889e-05,
+      "loss": 0.1947,
+      "step": 8360
+    },
+    {
+      "epoch": 5.978571428571429,
+      "grad_norm": 2.1099231243133545,
+      "learning_rate": 2.0851851851851852e-05,
+      "loss": 0.1847,
+      "step": 8370
+    },
+    {
+      "epoch": 5.985714285714286,
+      "grad_norm": 1.9181057214736938,
+      "learning_rate": 2.0814814814814813e-05,
+      "loss": 0.3513,
+      "step": 8380
+    },
+    {
+      "epoch": 5.992857142857143,
+      "grad_norm": 1.0576838254928589,
+      "learning_rate": 2.077777777777778e-05,
+      "loss": 0.2663,
+      "step": 8390
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.1283502578735352,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.2872,
+      "step": 8400
+    },
+    {
+      "epoch": 6.007142857142857,
+      "grad_norm": 0.7001394629478455,
+      "learning_rate": 2.0703703703703705e-05,
+      "loss": 0.277,
+      "step": 8410
+    },
+    {
+      "epoch": 6.014285714285714,
+      "grad_norm": 1.6374051570892334,
+      "learning_rate": 2.0666666666666666e-05,
+      "loss": 0.1849,
+      "step": 8420
+    },
+    {
+      "epoch": 6.021428571428571,
+      "grad_norm": 1.674914836883545,
+      "learning_rate": 2.0629629629629628e-05,
+      "loss": 0.1756,
+      "step": 8430
+    },
+    {
+      "epoch": 6.0285714285714285,
+      "grad_norm": 2.592038154602051,
+      "learning_rate": 2.0592592592592593e-05,
+      "loss": 0.3725,
+      "step": 8440
+    },
+    {
+      "epoch": 6.035714285714286,
+      "grad_norm": 2.942992925643921,
+      "learning_rate": 2.0555555555555555e-05,
+      "loss": 0.2529,
+      "step": 8450
+    },
+    {
+      "epoch": 6.042857142857143,
+      "grad_norm": 1.7580475807189941,
+      "learning_rate": 2.051851851851852e-05,
+      "loss": 0.1549,
+      "step": 8460
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.9032413959503174,
+      "learning_rate": 2.048148148148148e-05,
+      "loss": 0.2529,
+      "step": 8470
+    },
+    {
+      "epoch": 6.057142857142857,
+      "grad_norm": 1.7678323984146118,
+      "learning_rate": 2.0444444444444446e-05,
+      "loss": 0.1935,
+      "step": 8480
+    },
+    {
+      "epoch": 6.064285714285714,
+      "grad_norm": 1.7014952898025513,
+      "learning_rate": 2.0407407407407408e-05,
+      "loss": 0.1965,
+      "step": 8490
+    },
+    {
+      "epoch": 6.071428571428571,
+      "grad_norm": 2.053157091140747,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.2045,
+      "step": 8500
+    },
+    {
+      "epoch": 6.078571428571428,
+      "grad_norm": 2.448059320449829,
+      "learning_rate": 2.0333333333333334e-05,
+      "loss": 0.2275,
+      "step": 8510
+    },
+    {
+      "epoch": 6.085714285714285,
+      "grad_norm": 1.3505144119262695,
+      "learning_rate": 2.0296296296296296e-05,
+      "loss": 0.192,
+      "step": 8520
+    },
+    {
+      "epoch": 6.0928571428571425,
+      "grad_norm": 1.0717148780822754,
+      "learning_rate": 2.025925925925926e-05,
+      "loss": 0.3017,
+      "step": 8530
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 2.872880220413208,
+      "learning_rate": 2.0222222222222222e-05,
+      "loss": 0.2583,
+      "step": 8540
+    },
+    {
+      "epoch": 6.107142857142857,
+      "grad_norm": 1.559588074684143,
+      "learning_rate": 2.0185185185185187e-05,
+      "loss": 0.1557,
+      "step": 8550
+    },
+    {
+      "epoch": 6.114285714285714,
+      "grad_norm": 1.4375160932540894,
+      "learning_rate": 2.014814814814815e-05,
+      "loss": 0.1165,
+      "step": 8560
+    },
+    {
+      "epoch": 6.121428571428571,
+      "grad_norm": 1.1922268867492676,
+      "learning_rate": 2.011111111111111e-05,
+      "loss": 0.1995,
+      "step": 8570
+    },
+    {
+      "epoch": 6.128571428571428,
+      "grad_norm": 2.267056465148926,
+      "learning_rate": 2.0074074074074075e-05,
+      "loss": 0.2176,
+      "step": 8580
+    },
+    {
+      "epoch": 6.135714285714286,
+      "grad_norm": 1.5485496520996094,
+      "learning_rate": 2.0037037037037037e-05,
+      "loss": 0.206,
+      "step": 8590
+    },
+    {
+      "epoch": 6.142857142857143,
+      "grad_norm": 1.9538283348083496,
+      "learning_rate": 2e-05,
+      "loss": 0.3173,
+      "step": 8600
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 2.8216044902801514,
+      "learning_rate": 1.9962962962962963e-05,
+      "loss": 0.3077,
+      "step": 8610
+    },
+    {
+      "epoch": 6.1571428571428575,
+      "grad_norm": 2.5293240547180176,
+      "learning_rate": 1.9925925925925925e-05,
+      "loss": 0.2829,
+      "step": 8620
+    },
+    {
+      "epoch": 6.164285714285715,
+      "grad_norm": 1.7947183847427368,
+      "learning_rate": 1.988888888888889e-05,
+      "loss": 0.3212,
+      "step": 8630
+    },
+    {
+      "epoch": 6.171428571428572,
+      "grad_norm": 1.541588544845581,
+      "learning_rate": 1.985185185185185e-05,
+      "loss": 0.1985,
+      "step": 8640
+    },
+    {
+      "epoch": 6.178571428571429,
+      "grad_norm": 1.286007046699524,
+      "learning_rate": 1.9814814814814816e-05,
+      "loss": 0.279,
+      "step": 8650
+    },
+    {
+      "epoch": 6.185714285714286,
+      "grad_norm": 1.8692234754562378,
+      "learning_rate": 1.9777777777777778e-05,
+      "loss": 0.303,
+      "step": 8660
+    },
+    {
+      "epoch": 6.192857142857143,
+      "grad_norm": 1.7906513214111328,
+      "learning_rate": 1.9740740740740743e-05,
+      "loss": 0.218,
+      "step": 8670
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 2.0737709999084473,
+      "learning_rate": 1.9703703703703704e-05,
+      "loss": 0.1559,
+      "step": 8680
+    },
+    {
+      "epoch": 6.207142857142857,
+      "grad_norm": 1.8082749843597412,
+      "learning_rate": 1.9666666666666666e-05,
+      "loss": 0.2713,
+      "step": 8690
+    },
+    {
+      "epoch": 6.214285714285714,
+      "grad_norm": 1.8988617658615112,
+      "learning_rate": 1.962962962962963e-05,
+      "loss": 0.2362,
+      "step": 8700
+    },
+    {
+      "epoch": 6.2214285714285715,
+      "grad_norm": 0.8727281093597412,
+      "learning_rate": 1.9592592592592592e-05,
+      "loss": 0.2571,
+      "step": 8710
+    },
+    {
+      "epoch": 6.228571428571429,
+      "grad_norm": 1.0203776359558105,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 0.2884,
+      "step": 8720
+    },
+    {
+      "epoch": 6.235714285714286,
+      "grad_norm": 1.5776811838150024,
+      "learning_rate": 1.951851851851852e-05,
+      "loss": 0.3115,
+      "step": 8730
+    },
+    {
+      "epoch": 6.242857142857143,
+      "grad_norm": 2.1000545024871826,
+      "learning_rate": 1.948148148148148e-05,
+      "loss": 0.2936,
+      "step": 8740
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 1.991640329360962,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.2214,
+      "step": 8750
+    },
+    {
+      "epoch": 6.257142857142857,
+      "grad_norm": 3.8238637447357178,
+      "learning_rate": 1.9407407407407407e-05,
+      "loss": 0.2738,
+      "step": 8760
+    },
+    {
+      "epoch": 6.264285714285714,
+      "grad_norm": 2.818711042404175,
+      "learning_rate": 1.9370370370370372e-05,
+      "loss": 0.3107,
+      "step": 8770
+    },
+    {
+      "epoch": 6.271428571428571,
+      "grad_norm": 1.4565989971160889,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 0.1351,
+      "step": 8780
+    },
+    {
+      "epoch": 6.2785714285714285,
+      "grad_norm": 1.6833415031433105,
+      "learning_rate": 1.92962962962963e-05,
+      "loss": 0.3359,
+      "step": 8790
+    },
+    {
+      "epoch": 6.285714285714286,
+      "grad_norm": 3.662572145462036,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.2338,
+      "step": 8800
+    },
+    {
+      "epoch": 6.292857142857143,
+      "grad_norm": 1.9166165590286255,
+      "learning_rate": 1.922222222222222e-05,
+      "loss": 0.1908,
+      "step": 8810
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 2.665553331375122,
+      "learning_rate": 1.9185185185185186e-05,
+      "loss": 0.246,
+      "step": 8820
+    },
+    {
+      "epoch": 6.307142857142857,
+      "grad_norm": 1.601194143295288,
+      "learning_rate": 1.9148148148148148e-05,
+      "loss": 0.2392,
+      "step": 8830
+    },
+    {
+      "epoch": 6.314285714285714,
+      "grad_norm": 1.7382382154464722,
+      "learning_rate": 1.9111111111111113e-05,
+      "loss": 0.2919,
+      "step": 8840
+    },
+    {
+      "epoch": 6.321428571428571,
+      "grad_norm": 1.0822237730026245,
+      "learning_rate": 1.9074074074074075e-05,
+      "loss": 0.1179,
+      "step": 8850
+    },
+    {
+      "epoch": 6.328571428571428,
+      "grad_norm": 1.9691376686096191,
+      "learning_rate": 1.903703703703704e-05,
+      "loss": 0.3934,
+      "step": 8860
+    },
+    {
+      "epoch": 6.335714285714285,
+      "grad_norm": 0.8395001292228699,
+      "learning_rate": 1.9e-05,
+      "loss": 0.2004,
+      "step": 8870
+    },
+    {
+      "epoch": 6.3428571428571425,
+      "grad_norm": 1.6967720985412598,
+      "learning_rate": 1.8962962962962963e-05,
+      "loss": 0.204,
+      "step": 8880
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 1.2601035833358765,
+      "learning_rate": 1.8925925925925928e-05,
+      "loss": 0.2769,
+      "step": 8890
+    },
+    {
+      "epoch": 6.357142857142857,
+      "grad_norm": 1.560940146446228,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.1409,
+      "step": 8900
+    },
+    {
+      "epoch": 6.364285714285714,
+      "grad_norm": 1.645814061164856,
+      "learning_rate": 1.8851851851851854e-05,
+      "loss": 0.1914,
+      "step": 8910
+    },
+    {
+      "epoch": 6.371428571428572,
+      "grad_norm": 1.4886109828948975,
+      "learning_rate": 1.8814814814814816e-05,
+      "loss": 0.3517,
+      "step": 8920
+    },
+    {
+      "epoch": 6.378571428571428,
+      "grad_norm": 1.2002378702163696,
+      "learning_rate": 1.8777777777777777e-05,
+      "loss": 0.2346,
+      "step": 8930
+    },
+    {
+      "epoch": 6.385714285714286,
+      "grad_norm": 2.4492478370666504,
+      "learning_rate": 1.8740740740740742e-05,
+      "loss": 0.2104,
+      "step": 8940
+    },
+    {
+      "epoch": 6.392857142857143,
+      "grad_norm": 2.315610408782959,
+      "learning_rate": 1.8703703703703704e-05,
+      "loss": 0.2321,
+      "step": 8950
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.116260528564453,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.2092,
+      "step": 8960
+    },
+    {
+      "epoch": 6.4071428571428575,
+      "grad_norm": 1.7362505197525024,
+      "learning_rate": 1.862962962962963e-05,
+      "loss": 0.2598,
+      "step": 8970
+    },
+    {
+      "epoch": 6.414285714285715,
+      "grad_norm": 2.1754469871520996,
+      "learning_rate": 1.8592592592592595e-05,
+      "loss": 0.3035,
+      "step": 8980
+    },
+    {
+      "epoch": 6.421428571428572,
+      "grad_norm": 1.448285698890686,
+      "learning_rate": 1.8555555555555557e-05,
+      "loss": 0.227,
+      "step": 8990
+    },
+    {
+      "epoch": 6.428571428571429,
+      "grad_norm": 1.888242483139038,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.1618,
+      "step": 9000
+    },
+    {
+      "epoch": 6.428571428571429,
+      "eval_loss": 0.3434120714664459,
+      "eval_rouge1": 0.908,
+      "eval_rouge2": 0.8486,
+      "eval_rougeL": 0.9052,
+      "eval_runtime": 122.2937,
+      "eval_samples_per_second": 11.448,
+      "eval_steps_per_second": 5.724,
+      "step": 9000
+    },
+    {
+      "epoch": 6.435714285714286,
+      "grad_norm": 2.5552051067352295,
+      "learning_rate": 1.8481481481481483e-05,
+      "loss": 0.4376,
+      "step": 9010
+    },
+    {
+      "epoch": 6.442857142857143,
+      "grad_norm": 2.0973517894744873,
+      "learning_rate": 1.8444444444444445e-05,
+      "loss": 0.2163,
+      "step": 9020
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 1.3774244785308838,
+      "learning_rate": 1.840740740740741e-05,
+      "loss": 0.14,
+      "step": 9030
+    },
+    {
+      "epoch": 6.457142857142857,
+      "grad_norm": 0.8735131025314331,
+      "learning_rate": 1.837037037037037e-05,
+      "loss": 0.1848,
+      "step": 9040
+    },
+    {
+      "epoch": 6.464285714285714,
+      "grad_norm": 1.5088914632797241,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.2889,
+      "step": 9050
+    },
+    {
+      "epoch": 6.4714285714285715,
+      "grad_norm": 1.0688769817352295,
+      "learning_rate": 1.8296296296296298e-05,
+      "loss": 0.1895,
+      "step": 9060
+    },
+    {
+      "epoch": 6.478571428571429,
+      "grad_norm": 1.43760085105896,
+      "learning_rate": 1.825925925925926e-05,
+      "loss": 0.2997,
+      "step": 9070
+    },
+    {
+      "epoch": 6.485714285714286,
+      "grad_norm": 1.1168969869613647,
+      "learning_rate": 1.8222222222222224e-05,
+      "loss": 0.3544,
+      "step": 9080
+    },
+    {
+      "epoch": 6.492857142857143,
+      "grad_norm": 1.7139670848846436,
+      "learning_rate": 1.8185185185185186e-05,
+      "loss": 0.2108,
+      "step": 9090
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 1.2584503889083862,
+      "learning_rate": 1.814814814814815e-05,
+      "loss": 0.2791,
+      "step": 9100
+    },
+    {
+      "epoch": 6.507142857142857,
+      "grad_norm": 1.4440019130706787,
+      "learning_rate": 1.8111111111111112e-05,
+      "loss": 0.3745,
+      "step": 9110
+    },
+    {
+      "epoch": 6.514285714285714,
+      "grad_norm": 2.3828232288360596,
+      "learning_rate": 1.8074074074074074e-05,
+      "loss": 0.2159,
+      "step": 9120
+    },
+    {
+      "epoch": 6.521428571428571,
+      "grad_norm": 2.6553053855895996,
+      "learning_rate": 1.803703703703704e-05,
+      "loss": 0.3051,
+      "step": 9130
+    },
+    {
+      "epoch": 6.5285714285714285,
+      "grad_norm": 2.0669426918029785,
+      "learning_rate": 1.8e-05,
+      "loss": 0.2166,
+      "step": 9140
+    },
+    {
+      "epoch": 6.535714285714286,
+      "grad_norm": 1.4676064252853394,
+      "learning_rate": 1.7962962962962965e-05,
+      "loss": 0.2393,
+      "step": 9150
+    },
+    {
+      "epoch": 6.542857142857143,
+      "grad_norm": 1.5158963203430176,
+      "learning_rate": 1.7925925925925927e-05,
+      "loss": 0.2821,
+      "step": 9160
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 1.438550591468811,
+      "learning_rate": 1.788888888888889e-05,
+      "loss": 0.229,
+      "step": 9170
+    },
+    {
+      "epoch": 6.557142857142857,
+      "grad_norm": 2.2161788940429688,
+      "learning_rate": 1.7851851851851853e-05,
+      "loss": 0.3705,
+      "step": 9180
+    },
+    {
+      "epoch": 6.564285714285714,
+      "grad_norm": 1.472321629524231,
+      "learning_rate": 1.7814814814814815e-05,
+      "loss": 0.3977,
+      "step": 9190
+    },
+    {
+      "epoch": 6.571428571428571,
+      "grad_norm": 1.957033395767212,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.2431,
+      "step": 9200
+    },
+    {
+      "epoch": 6.578571428571428,
+      "grad_norm": 3.070905923843384,
+      "learning_rate": 1.774074074074074e-05,
+      "loss": 0.2676,
+      "step": 9210
+    },
+    {
+      "epoch": 6.585714285714285,
+      "grad_norm": 2.240701198577881,
+      "learning_rate": 1.7703703703703706e-05,
+      "loss": 0.2346,
+      "step": 9220
+    },
+    {
+      "epoch": 6.5928571428571425,
+      "grad_norm": 1.2726478576660156,
+      "learning_rate": 1.7666666666666668e-05,
+      "loss": 0.2624,
+      "step": 9230
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 2.543856382369995,
+      "learning_rate": 1.762962962962963e-05,
+      "loss": 0.3137,
+      "step": 9240
+    },
+    {
+      "epoch": 6.607142857142857,
+      "grad_norm": 2.1688966751098633,
+      "learning_rate": 1.7592592592592595e-05,
+      "loss": 0.3366,
+      "step": 9250
+    },
+    {
+      "epoch": 6.614285714285714,
+      "grad_norm": 1.9013522863388062,
+      "learning_rate": 1.7555555555555556e-05,
+      "loss": 0.1759,
+      "step": 9260
+    },
+    {
+      "epoch": 6.621428571428572,
+      "grad_norm": 2.7567338943481445,
+      "learning_rate": 1.751851851851852e-05,
+      "loss": 0.2615,
+      "step": 9270
+    },
+    {
+      "epoch": 6.628571428571428,
+      "grad_norm": 2.530351161956787,
+      "learning_rate": 1.7481481481481483e-05,
+      "loss": 0.363,
+      "step": 9280
+    },
+    {
+      "epoch": 6.635714285714286,
+      "grad_norm": 3.0051562786102295,
+      "learning_rate": 1.7444444444444448e-05,
+      "loss": 0.2155,
+      "step": 9290
+    },
+    {
+      "epoch": 6.642857142857143,
+      "grad_norm": 1.6199374198913574,
+      "learning_rate": 1.740740740740741e-05,
+      "loss": 0.1943,
+      "step": 9300
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 2.2254199981689453,
+      "learning_rate": 1.737037037037037e-05,
+      "loss": 0.2086,
+      "step": 9310
+    },
+    {
+      "epoch": 6.6571428571428575,
+      "grad_norm": 1.4565106630325317,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.2113,
+      "step": 9320
+    },
+    {
+      "epoch": 6.664285714285715,
+      "grad_norm": 1.8667312860488892,
+      "learning_rate": 1.7296296296296297e-05,
+      "loss": 0.1719,
+      "step": 9330
+    },
+    {
+      "epoch": 6.671428571428572,
+      "grad_norm": 2.0462963581085205,
+      "learning_rate": 1.7259259259259262e-05,
+      "loss": 0.2307,
+      "step": 9340
+    },
+    {
+      "epoch": 6.678571428571429,
+      "grad_norm": 1.5114613771438599,
+      "learning_rate": 1.7222222222222224e-05,
+      "loss": 0.2629,
+      "step": 9350
+    },
+    {
+      "epoch": 6.685714285714286,
+      "grad_norm": 1.8743935823440552,
+      "learning_rate": 1.7185185185185185e-05,
+      "loss": 0.2656,
+      "step": 9360
+    },
+    {
+      "epoch": 6.692857142857143,
+      "grad_norm": 1.6508034467697144,
+      "learning_rate": 1.714814814814815e-05,
+      "loss": 0.2971,
+      "step": 9370
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.4109563827514648,
+      "learning_rate": 1.7111111111111112e-05,
+      "loss": 0.3155,
+      "step": 9380
+    },
+    {
+      "epoch": 6.707142857142857,
+      "grad_norm": 1.9742975234985352,
+      "learning_rate": 1.7074074074074077e-05,
+      "loss": 0.2858,
+      "step": 9390
+    },
+    {
+      "epoch": 6.714285714285714,
+      "grad_norm": 0.8593278527259827,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.2484,
+      "step": 9400
+    },
+    {
+      "epoch": 6.7214285714285715,
+      "grad_norm": 1.8331007957458496,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.2763,
+      "step": 9410
+    },
+    {
+      "epoch": 6.728571428571429,
+      "grad_norm": 2.0606274604797363,
+      "learning_rate": 1.6962962962962965e-05,
+      "loss": 0.2016,
+      "step": 9420
+    },
+    {
+      "epoch": 6.735714285714286,
+      "grad_norm": 1.244935154914856,
+      "learning_rate": 1.6925925925925926e-05,
+      "loss": 0.2161,
+      "step": 9430
+    },
+    {
+      "epoch": 6.742857142857143,
+      "grad_norm": 2.0855889320373535,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 0.1961,
+      "step": 9440
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 2.203310012817383,
+      "learning_rate": 1.6851851851851853e-05,
+      "loss": 0.1886,
+      "step": 9450
+    },
+    {
+      "epoch": 6.757142857142857,
+      "grad_norm": 2.1254501342773438,
+      "learning_rate": 1.6814814814814818e-05,
+      "loss": 0.2824,
+      "step": 9460
+    },
+    {
+      "epoch": 6.764285714285714,
+      "grad_norm": 1.498728632926941,
+      "learning_rate": 1.677777777777778e-05,
+      "loss": 0.2848,
+      "step": 9470
+    },
+    {
+      "epoch": 6.771428571428571,
+      "grad_norm": 2.6205763816833496,
+      "learning_rate": 1.674074074074074e-05,
+      "loss": 0.2728,
+      "step": 9480
+    },
+    {
+      "epoch": 6.7785714285714285,
+      "grad_norm": 1.6262216567993164,
+      "learning_rate": 1.6703703703703706e-05,
+      "loss": 0.4216,
+      "step": 9490
+    },
+    {
+      "epoch": 6.785714285714286,
+      "grad_norm": 3.074489116668701,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2084,
+      "step": 9500
+    },
+    {
+      "epoch": 6.792857142857143,
+      "grad_norm": 1.8158230781555176,
+      "learning_rate": 1.6629629629629632e-05,
+      "loss": 0.1794,
+      "step": 9510
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.069397449493408,
+      "learning_rate": 1.6592592592592594e-05,
+      "loss": 0.2363,
+      "step": 9520
+    },
+    {
+      "epoch": 6.807142857142857,
+      "grad_norm": 1.8637501001358032,
+      "learning_rate": 1.655555555555556e-05,
+      "loss": 0.2203,
+      "step": 9530
+    },
+    {
+      "epoch": 6.814285714285714,
+      "grad_norm": 2.043314218521118,
+      "learning_rate": 1.651851851851852e-05,
+      "loss": 0.2267,
+      "step": 9540
+    },
+    {
+      "epoch": 6.821428571428571,
+      "grad_norm": 2.8327081203460693,
+      "learning_rate": 1.6481481481481482e-05,
+      "loss": 0.2793,
+      "step": 9550
+    },
+    {
+      "epoch": 6.828571428571428,
+      "grad_norm": 2.3297407627105713,
+      "learning_rate": 1.6444444444444447e-05,
+      "loss": 0.2349,
+      "step": 9560
+    },
+    {
+      "epoch": 6.835714285714285,
+      "grad_norm": 0.9220748543739319,
+      "learning_rate": 1.640740740740741e-05,
+      "loss": 0.1966,
+      "step": 9570
+    },
+    {
+      "epoch": 6.8428571428571425,
+      "grad_norm": 1.5935183763504028,
+      "learning_rate": 1.6370370370370374e-05,
+      "loss": 0.3217,
+      "step": 9580
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.9305605292320251,
+      "learning_rate": 1.6333333333333335e-05,
+      "loss": 0.1446,
+      "step": 9590
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 2.0719094276428223,
+      "learning_rate": 1.62962962962963e-05,
+      "loss": 0.2195,
+      "step": 9600
+    },
+    {
+      "epoch": 6.864285714285714,
+      "grad_norm": 1.9230345487594604,
+      "learning_rate": 1.625925925925926e-05,
+      "loss": 0.2031,
+      "step": 9610
+    },
+    {
+      "epoch": 6.871428571428572,
+      "grad_norm": 1.7897018194198608,
+      "learning_rate": 1.6222222222222223e-05,
+      "loss": 0.1728,
+      "step": 9620
+    },
+    {
+      "epoch": 6.878571428571428,
+      "grad_norm": 2.4588770866394043,
+      "learning_rate": 1.6185185185185188e-05,
+      "loss": 0.3253,
+      "step": 9630
+    },
+    {
+      "epoch": 6.885714285714286,
+      "grad_norm": 1.2495237588882446,
+      "learning_rate": 1.614814814814815e-05,
+      "loss": 0.3539,
+      "step": 9640
+    },
+    {
+      "epoch": 6.892857142857143,
+      "grad_norm": 3.161078453063965,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 0.3598,
+      "step": 9650
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.9474009275436401,
+      "learning_rate": 1.6074074074074076e-05,
+      "loss": 0.1385,
+      "step": 9660
+    },
+    {
+      "epoch": 6.9071428571428575,
+      "grad_norm": 1.9687261581420898,
+      "learning_rate": 1.6037037037037038e-05,
+      "loss": 0.2375,
+      "step": 9670
+    },
+    {
+      "epoch": 6.914285714285715,
+      "grad_norm": 1.87405264377594,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.3374,
+      "step": 9680
+    },
+    {
+      "epoch": 6.921428571428572,
+      "grad_norm": 1.1928725242614746,
+      "learning_rate": 1.5962962962962964e-05,
+      "loss": 0.197,
+      "step": 9690
+    },
+    {
+      "epoch": 6.928571428571429,
+      "grad_norm": 1.6434850692749023,
+      "learning_rate": 1.5925925925925926e-05,
+      "loss": 0.2757,
+      "step": 9700
+    },
+    {
+      "epoch": 6.935714285714286,
+      "grad_norm": 1.1469305753707886,
+      "learning_rate": 1.588888888888889e-05,
+      "loss": 0.1793,
+      "step": 9710
+    },
+    {
+      "epoch": 6.942857142857143,
+      "grad_norm": 2.436051368713379,
+      "learning_rate": 1.5851851851851852e-05,
+      "loss": 0.3178,
+      "step": 9720
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 1.8960529565811157,
+      "learning_rate": 1.5814814814814817e-05,
+      "loss": 0.2182,
+      "step": 9730
+    },
+    {
+      "epoch": 6.957142857142857,
+      "grad_norm": 1.755922794342041,
+      "learning_rate": 1.577777777777778e-05,
+      "loss": 0.3239,
+      "step": 9740
+    },
+    {
+      "epoch": 6.964285714285714,
+      "grad_norm": 2.202697515487671,
+      "learning_rate": 1.574074074074074e-05,
+      "loss": 0.2116,
+      "step": 9750
+    },
+    {
+      "epoch": 6.9714285714285715,
+      "grad_norm": 1.4491599798202515,
+      "learning_rate": 1.5703703703703705e-05,
+      "loss": 0.2329,
+      "step": 9760
+    },
+    {
+      "epoch": 6.978571428571429,
+      "grad_norm": 9.212343215942383,
+      "learning_rate": 1.5666666666666667e-05,
+      "loss": 0.2573,
+      "step": 9770
+    },
+    {
+      "epoch": 6.985714285714286,
+      "grad_norm": 1.2211856842041016,
+      "learning_rate": 1.5629629629629632e-05,
+      "loss": 0.2737,
+      "step": 9780
+    },
+    {
+      "epoch": 6.992857142857143,
+      "grad_norm": 1.59877347946167,
+      "learning_rate": 1.5592592592592593e-05,
+      "loss": 0.284,
+      "step": 9790
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.295945882797241,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.2076,
+      "step": 9800
+    },
+    {
+      "epoch": 7.007142857142857,
+      "grad_norm": 1.4388489723205566,
+      "learning_rate": 1.551851851851852e-05,
+      "loss": 0.2225,
+      "step": 9810
+    },
+    {
+      "epoch": 7.014285714285714,
+      "grad_norm": 1.9146931171417236,
+      "learning_rate": 1.548148148148148e-05,
+      "loss": 0.2917,
+      "step": 9820
+    },
+    {
+      "epoch": 7.021428571428571,
+      "grad_norm": 1.0212804079055786,
+      "learning_rate": 1.5444444444444446e-05,
+      "loss": 0.1537,
+      "step": 9830
+    },
+    {
+      "epoch": 7.0285714285714285,
+      "grad_norm": 2.146648645401001,
+      "learning_rate": 1.5407407407407408e-05,
+      "loss": 0.178,
+      "step": 9840
+    },
+    {
+      "epoch": 7.035714285714286,
+      "grad_norm": 2.4515628814697266,
+      "learning_rate": 1.537037037037037e-05,
+      "loss": 0.3043,
+      "step": 9850
+    },
+    {
+      "epoch": 7.042857142857143,
+      "grad_norm": 1.6906862258911133,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.2787,
+      "step": 9860
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 2.2019400596618652,
+      "learning_rate": 1.5296296296296296e-05,
+      "loss": 0.3236,
+      "step": 9870
+    },
+    {
+      "epoch": 7.057142857142857,
+      "grad_norm": 1.3307303190231323,
+      "learning_rate": 1.5259259259259258e-05,
+      "loss": 0.1875,
+      "step": 9880
+    },
+    {
+      "epoch": 7.064285714285714,
+      "grad_norm": 1.7358342409133911,
+      "learning_rate": 1.5222222222222224e-05,
+      "loss": 0.2149,
+      "step": 9890
+    },
+    {
+      "epoch": 7.071428571428571,
+      "grad_norm": 2.0298547744750977,
+      "learning_rate": 1.5185185185185186e-05,
+      "loss": 0.1876,
+      "step": 9900
+    },
+    {
+      "epoch": 7.078571428571428,
+      "grad_norm": 2.375779151916504,
+      "learning_rate": 1.514814814814815e-05,
+      "loss": 0.2289,
+      "step": 9910
+    },
+    {
+      "epoch": 7.085714285714285,
+      "grad_norm": 1.856911540031433,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 0.2029,
+      "step": 9920
+    },
+    {
+      "epoch": 7.0928571428571425,
+      "grad_norm": 1.1523020267486572,
+      "learning_rate": 1.5074074074074074e-05,
+      "loss": 0.1753,
+      "step": 9930
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 1.4677330255508423,
+      "learning_rate": 1.5037037037037039e-05,
+      "loss": 0.2256,
+      "step": 9940
+    },
+    {
+      "epoch": 7.107142857142857,
+      "grad_norm": 1.0742135047912598,
+      "learning_rate": 1.5e-05,
+      "loss": 0.3844,
+      "step": 9950
+    },
+    {
+      "epoch": 7.114285714285714,
+      "grad_norm": 1.4122258424758911,
+      "learning_rate": 1.4962962962962965e-05,
+      "loss": 0.1498,
+      "step": 9960
+    },
+    {
+      "epoch": 7.121428571428571,
+      "grad_norm": 1.9363057613372803,
+      "learning_rate": 1.4925925925925927e-05,
+      "loss": 0.2721,
+      "step": 9970
+    },
+    {
+      "epoch": 7.128571428571428,
+      "grad_norm": 0.7882018685340881,
+      "learning_rate": 1.4888888888888888e-05,
+      "loss": 0.2363,
+      "step": 9980
+    },
+    {
+      "epoch": 7.135714285714286,
+      "grad_norm": 1.8561784029006958,
+      "learning_rate": 1.4851851851851853e-05,
+      "loss": 0.2875,
+      "step": 9990
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 1.9597991704940796,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.1984,
+      "step": 10000
+    },
+    {
+      "epoch": 7.142857142857143,
+      "eval_loss": 0.34164437651634216,
+      "eval_rouge1": 0.9089,
+      "eval_rouge2": 0.8501,
+      "eval_rougeL": 0.9063,
+      "eval_runtime": 122.2974,
+      "eval_samples_per_second": 11.448,
+      "eval_steps_per_second": 5.724,
+      "step": 10000
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 2.1984336376190186,
+      "learning_rate": 1.477777777777778e-05,
+      "loss": 0.2409,
+      "step": 10010
+    },
+    {
+      "epoch": 7.1571428571428575,
+      "grad_norm": 1.1259089708328247,
+      "learning_rate": 1.4740740740740741e-05,
+      "loss": 0.1386,
+      "step": 10020
+    },
+    {
+      "epoch": 7.164285714285715,
+      "grad_norm": 2.6497113704681396,
+      "learning_rate": 1.4703703703703706e-05,
+      "loss": 0.2999,
+      "step": 10030
+    },
+    {
+      "epoch": 7.171428571428572,
+      "grad_norm": 2.7574968338012695,
+      "learning_rate": 1.4666666666666668e-05,
+      "loss": 0.2064,
+      "step": 10040
+    },
+    {
+      "epoch": 7.178571428571429,
+      "grad_norm": 2.4020519256591797,
+      "learning_rate": 1.462962962962963e-05,
+      "loss": 0.2539,
+      "step": 10050
+    },
+    {
+      "epoch": 7.185714285714286,
+      "grad_norm": 1.8728407621383667,
+      "learning_rate": 1.4592592592592594e-05,
+      "loss": 0.2264,
+      "step": 10060
+    },
+    {
+      "epoch": 7.192857142857143,
+      "grad_norm": 3.187389612197876,
+      "learning_rate": 1.4555555555555556e-05,
+      "loss": 0.2175,
+      "step": 10070
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 1.6992945671081543,
+      "learning_rate": 1.4518518518518521e-05,
+      "loss": 0.1752,
+      "step": 10080
+    },
+    {
+      "epoch": 7.207142857142857,
+      "grad_norm": 1.2373261451721191,
+      "learning_rate": 1.4481481481481483e-05,
+      "loss": 0.1712,
+      "step": 10090
+    },
+    {
+      "epoch": 7.214285714285714,
+      "grad_norm": 1.3986244201660156,
+      "learning_rate": 1.4444444444444444e-05,
+      "loss": 0.1727,
+      "step": 10100
+    },
+    {
+      "epoch": 7.2214285714285715,
+      "grad_norm": 1.5018147230148315,
+      "learning_rate": 1.4407407407407409e-05,
+      "loss": 0.2309,
+      "step": 10110
+    },
+    {
+      "epoch": 7.228571428571429,
+      "grad_norm": 1.8186851739883423,
+      "learning_rate": 1.437037037037037e-05,
+      "loss": 0.2741,
+      "step": 10120
+    },
+    {
+      "epoch": 7.235714285714286,
+      "grad_norm": 2.8224360942840576,
+      "learning_rate": 1.4333333333333334e-05,
+      "loss": 0.2244,
+      "step": 10130
+    },
+    {
+      "epoch": 7.242857142857143,
+      "grad_norm": 1.900585412979126,
+      "learning_rate": 1.4296296296296297e-05,
+      "loss": 0.2527,
+      "step": 10140
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 2.4210896492004395,
+      "learning_rate": 1.425925925925926e-05,
+      "loss": 0.3832,
+      "step": 10150
+    },
+    {
+      "epoch": 7.257142857142857,
+      "grad_norm": 1.2783209085464478,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 0.1756,
+      "step": 10160
+    },
+    {
+      "epoch": 7.264285714285714,
+      "grad_norm": 1.4387212991714478,
+      "learning_rate": 1.4185185185185185e-05,
+      "loss": 0.2567,
+      "step": 10170
+    },
+    {
+      "epoch": 7.271428571428571,
+      "grad_norm": 2.861311435699463,
+      "learning_rate": 1.4148148148148148e-05,
+      "loss": 0.3683,
+      "step": 10180
+    },
+    {
+      "epoch": 7.2785714285714285,
+      "grad_norm": 0.8701191544532776,
+      "learning_rate": 1.4111111111111112e-05,
+      "loss": 0.2229,
+      "step": 10190
+    },
+    {
+      "epoch": 7.285714285714286,
+      "grad_norm": 2.103231430053711,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.1652,
+      "step": 10200
+    },
+    {
+      "epoch": 7.292857142857143,
+      "grad_norm": 3.0958895683288574,
+      "learning_rate": 1.4037037037037038e-05,
+      "loss": 0.3048,
+      "step": 10210
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 1.0370267629623413,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.243,
+      "step": 10220
+    },
+    {
+      "epoch": 7.307142857142857,
+      "grad_norm": 1.334799885749817,
+      "learning_rate": 1.3962962962962963e-05,
+      "loss": 0.2242,
+      "step": 10230
+    },
+    {
+      "epoch": 7.314285714285714,
+      "grad_norm": 1.797135829925537,
+      "learning_rate": 1.3925925925925926e-05,
+      "loss": 0.1916,
+      "step": 10240
+    },
+    {
+      "epoch": 7.321428571428571,
+      "grad_norm": 0.7588611841201782,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.2548,
+      "step": 10250
+    },
+    {
+      "epoch": 7.328571428571428,
+      "grad_norm": 1.9136377573013306,
+      "learning_rate": 1.3851851851851853e-05,
+      "loss": 0.2373,
+      "step": 10260
+    },
+    {
+      "epoch": 7.335714285714285,
+      "grad_norm": 1.326635479927063,
+      "learning_rate": 1.3814814814814816e-05,
+      "loss": 0.2597,
+      "step": 10270
+    },
+    {
+      "epoch": 7.3428571428571425,
+      "grad_norm": 2.400609016418457,
+      "learning_rate": 1.3777777777777778e-05,
+      "loss": 0.2648,
+      "step": 10280
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.9898678064346313,
+      "learning_rate": 1.3740740740740741e-05,
+      "loss": 0.2304,
+      "step": 10290
+    },
+    {
+      "epoch": 7.357142857142857,
+      "grad_norm": 0.7826656103134155,
+      "learning_rate": 1.3703703703703704e-05,
+      "loss": 0.1209,
+      "step": 10300
+    },
+    {
+      "epoch": 7.364285714285714,
+      "grad_norm": 1.083044409751892,
+      "learning_rate": 1.3666666666666666e-05,
+      "loss": 0.2026,
+      "step": 10310
+    },
+    {
+      "epoch": 7.371428571428572,
+      "grad_norm": 1.283219814300537,
+      "learning_rate": 1.362962962962963e-05,
+      "loss": 0.2018,
+      "step": 10320
+    },
+    {
+      "epoch": 7.378571428571428,
+      "grad_norm": 1.9941823482513428,
+      "learning_rate": 1.3592592592592592e-05,
+      "loss": 0.2469,
+      "step": 10330
+    },
+    {
+      "epoch": 7.385714285714286,
+      "grad_norm": 2.553957462310791,
+      "learning_rate": 1.3555555555555557e-05,
+      "loss": 0.2661,
+      "step": 10340
+    },
+    {
+      "epoch": 7.392857142857143,
+      "grad_norm": 1.657182216644287,
+      "learning_rate": 1.3518518518518519e-05,
+      "loss": 0.2762,
+      "step": 10350
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 1.6704496145248413,
+      "learning_rate": 1.348148148148148e-05,
+      "loss": 0.222,
+      "step": 10360
+    },
+    {
+      "epoch": 7.4071428571428575,
+      "grad_norm": 1.338329792022705,
+      "learning_rate": 1.3444444444444445e-05,
+      "loss": 0.2658,
+      "step": 10370
+    },
+    {
+      "epoch": 7.414285714285715,
+      "grad_norm": 1.9741250276565552,
+      "learning_rate": 1.3407407407407407e-05,
+      "loss": 0.2596,
+      "step": 10380
+    },
+    {
+      "epoch": 7.421428571428572,
+      "grad_norm": 2.523958444595337,
+      "learning_rate": 1.3370370370370372e-05,
+      "loss": 0.1553,
+      "step": 10390
+    },
+    {
+      "epoch": 7.428571428571429,
+      "grad_norm": 2.260690450668335,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.2942,
+      "step": 10400
+    },
+    {
+      "epoch": 7.435714285714286,
+      "grad_norm": 2.441620349884033,
+      "learning_rate": 1.3296296296296295e-05,
+      "loss": 0.2451,
+      "step": 10410
+    },
+    {
+      "epoch": 7.442857142857143,
+      "grad_norm": 0.7054124474525452,
+      "learning_rate": 1.325925925925926e-05,
+      "loss": 0.1862,
+      "step": 10420
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 1.6281330585479736,
+      "learning_rate": 1.3222222222222221e-05,
+      "loss": 0.1714,
+      "step": 10430
+    },
+    {
+      "epoch": 7.457142857142857,
+      "grad_norm": 1.738685965538025,
+      "learning_rate": 1.3185185185185186e-05,
+      "loss": 0.2052,
+      "step": 10440
+    },
+    {
+      "epoch": 7.464285714285714,
+      "grad_norm": 1.9982494115829468,
+      "learning_rate": 1.3148148148148148e-05,
+      "loss": 0.2964,
+      "step": 10450
+    },
+    {
+      "epoch": 7.4714285714285715,
+      "grad_norm": 1.0081127882003784,
+      "learning_rate": 1.3111111111111113e-05,
+      "loss": 0.1956,
+      "step": 10460
+    },
+    {
+      "epoch": 7.478571428571429,
+      "grad_norm": 1.5927938222885132,
+      "learning_rate": 1.3074074074074074e-05,
+      "loss": 0.236,
+      "step": 10470
+    },
+    {
+      "epoch": 7.485714285714286,
+      "grad_norm": 1.7959505319595337,
+      "learning_rate": 1.3037037037037036e-05,
+      "loss": 0.2462,
+      "step": 10480
+    },
+    {
+      "epoch": 7.492857142857143,
+      "grad_norm": 0.9230768084526062,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.177,
+      "step": 10490
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 1.8714969158172607,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.2224,
+      "step": 10500
+    },
+    {
+      "epoch": 7.507142857142857,
+      "grad_norm": 2.809420585632324,
+      "learning_rate": 1.2925925925925927e-05,
+      "loss": 0.2214,
+      "step": 10510
+    },
+    {
+      "epoch": 7.514285714285714,
+      "grad_norm": 2.2183682918548584,
+      "learning_rate": 1.2888888888888889e-05,
+      "loss": 0.2528,
+      "step": 10520
+    },
+    {
+      "epoch": 7.521428571428571,
+      "grad_norm": 2.108675479888916,
+      "learning_rate": 1.2851851851851854e-05,
+      "loss": 0.2593,
+      "step": 10530
+    },
+    {
+      "epoch": 7.5285714285714285,
+      "grad_norm": 1.9557310342788696,
+      "learning_rate": 1.2814814814814815e-05,
+      "loss": 0.1875,
+      "step": 10540
+    },
+    {
+      "epoch": 7.535714285714286,
+      "grad_norm": 2.3365464210510254,
+      "learning_rate": 1.2777777777777777e-05,
+      "loss": 0.2232,
+      "step": 10550
+    },
+    {
+      "epoch": 7.542857142857143,
+      "grad_norm": 1.4713023900985718,
+      "learning_rate": 1.2740740740740742e-05,
+      "loss": 0.2592,
+      "step": 10560
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 2.5241403579711914,
+      "learning_rate": 1.2703703703703704e-05,
+      "loss": 0.2632,
+      "step": 10570
+    },
+    {
+      "epoch": 7.557142857142857,
+      "grad_norm": 2.726618528366089,
+      "learning_rate": 1.2666666666666668e-05,
+      "loss": 0.2068,
+      "step": 10580
+    },
+    {
+      "epoch": 7.564285714285714,
+      "grad_norm": 1.2947627305984497,
+      "learning_rate": 1.262962962962963e-05,
+      "loss": 0.1846,
+      "step": 10590
+    },
+    {
+      "epoch": 7.571428571428571,
+      "grad_norm": 1.4739402532577515,
+      "learning_rate": 1.2592592592592592e-05,
+      "loss": 0.1445,
+      "step": 10600
+    },
+    {
+      "epoch": 7.578571428571428,
+      "grad_norm": 1.7607239484786987,
+      "learning_rate": 1.2555555555555557e-05,
+      "loss": 0.2087,
+      "step": 10610
+    },
+    {
+      "epoch": 7.585714285714285,
+      "grad_norm": 1.510556697845459,
+      "learning_rate": 1.2518518518518518e-05,
+      "loss": 0.2356,
+      "step": 10620
+    },
+    {
+      "epoch": 7.5928571428571425,
+      "grad_norm": 1.4189872741699219,
+      "learning_rate": 1.2481481481481481e-05,
+      "loss": 0.2343,
+      "step": 10630
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 3.209477424621582,
+      "learning_rate": 1.2444444444444445e-05,
+      "loss": 0.2131,
+      "step": 10640
+    },
+    {
+      "epoch": 7.607142857142857,
+      "grad_norm": 2.026301145553589,
+      "learning_rate": 1.2407407407407408e-05,
+      "loss": 0.3643,
+      "step": 10650
+    },
+    {
+      "epoch": 7.614285714285714,
+      "grad_norm": 2.355459451675415,
+      "learning_rate": 1.2370370370370371e-05,
+      "loss": 0.2382,
+      "step": 10660
+    },
+    {
+      "epoch": 7.621428571428572,
+      "grad_norm": 1.6867364645004272,
+      "learning_rate": 1.2333333333333334e-05,
+      "loss": 0.1886,
+      "step": 10670
+    },
+    {
+      "epoch": 7.628571428571428,
+      "grad_norm": 1.863373041152954,
+      "learning_rate": 1.2296296296296298e-05,
+      "loss": 0.2842,
+      "step": 10680
+    },
+    {
+      "epoch": 7.635714285714286,
+      "grad_norm": 1.4037106037139893,
+      "learning_rate": 1.225925925925926e-05,
+      "loss": 0.297,
+      "step": 10690
+    },
+    {
+      "epoch": 7.642857142857143,
+      "grad_norm": 1.2220287322998047,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 0.2126,
+      "step": 10700
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 1.796430230140686,
+      "learning_rate": 1.2185185185185186e-05,
+      "loss": 0.2494,
+      "step": 10710
+    },
+    {
+      "epoch": 7.6571428571428575,
+      "grad_norm": 2.7764432430267334,
+      "learning_rate": 1.2148148148148149e-05,
+      "loss": 0.3102,
+      "step": 10720
+    },
+    {
+      "epoch": 7.664285714285715,
+      "grad_norm": 2.2875261306762695,
+      "learning_rate": 1.2111111111111112e-05,
+      "loss": 0.2171,
+      "step": 10730
+    },
+    {
+      "epoch": 7.671428571428572,
+      "grad_norm": 1.3017419576644897,
+      "learning_rate": 1.2074074074074075e-05,
+      "loss": 0.2095,
+      "step": 10740
+    },
+    {
+      "epoch": 7.678571428571429,
+      "grad_norm": 1.5423152446746826,
+      "learning_rate": 1.2037037037037037e-05,
+      "loss": 0.3183,
+      "step": 10750
+    },
+    {
+      "epoch": 7.685714285714286,
+      "grad_norm": 2.0346460342407227,
+      "learning_rate": 1.2e-05,
+      "loss": 0.1669,
+      "step": 10760
+    },
+    {
+      "epoch": 7.692857142857143,
+      "grad_norm": 1.178389549255371,
+      "learning_rate": 1.1962962962962964e-05,
+      "loss": 0.3195,
+      "step": 10770
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 1.3902812004089355,
+      "learning_rate": 1.1925925925925927e-05,
+      "loss": 0.2166,
+      "step": 10780
+    },
+    {
+      "epoch": 7.707142857142857,
+      "grad_norm": 2.8894922733306885,
+      "learning_rate": 1.188888888888889e-05,
+      "loss": 0.2216,
+      "step": 10790
+    },
+    {
+      "epoch": 7.714285714285714,
+      "grad_norm": 2.77864670753479,
+      "learning_rate": 1.1851851851851853e-05,
+      "loss": 0.224,
+      "step": 10800
+    },
+    {
+      "epoch": 7.7214285714285715,
+      "grad_norm": 1.370814323425293,
+      "learning_rate": 1.1814814814814817e-05,
+      "loss": 0.2277,
+      "step": 10810
+    },
+    {
+      "epoch": 7.728571428571429,
+      "grad_norm": 1.3543068170547485,
+      "learning_rate": 1.1777777777777778e-05,
+      "loss": 0.19,
+      "step": 10820
+    },
+    {
+      "epoch": 7.735714285714286,
+      "grad_norm": 2.4707486629486084,
+      "learning_rate": 1.1740740740740741e-05,
+      "loss": 0.1951,
+      "step": 10830
+    },
+    {
+      "epoch": 7.742857142857143,
+      "grad_norm": 2.284876823425293,
+      "learning_rate": 1.1703703703703705e-05,
+      "loss": 0.2206,
+      "step": 10840
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 1.1018098592758179,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.1386,
+      "step": 10850
+    },
+    {
+      "epoch": 7.757142857142857,
+      "grad_norm": 1.9555597305297852,
+      "learning_rate": 1.1629629629629631e-05,
+      "loss": 0.1645,
+      "step": 10860
+    },
+    {
+      "epoch": 7.764285714285714,
+      "grad_norm": 1.1327388286590576,
+      "learning_rate": 1.1592592592592594e-05,
+      "loss": 0.1654,
+      "step": 10870
+    },
+    {
+      "epoch": 7.771428571428571,
+      "grad_norm": 1.6210479736328125,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 0.1937,
+      "step": 10880
+    },
+    {
+      "epoch": 7.7785714285714285,
+      "grad_norm": 1.7142146825790405,
+      "learning_rate": 1.151851851851852e-05,
+      "loss": 0.1722,
+      "step": 10890
+    },
+    {
+      "epoch": 7.785714285714286,
+      "grad_norm": 3.529614210128784,
+      "learning_rate": 1.1481481481481482e-05,
+      "loss": 0.2414,
+      "step": 10900
+    },
+    {
+      "epoch": 7.792857142857143,
+      "grad_norm": 1.6331572532653809,
+      "learning_rate": 1.1444444444444446e-05,
+      "loss": 0.1844,
+      "step": 10910
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 2.7431063652038574,
+      "learning_rate": 1.1407407407407409e-05,
+      "loss": 0.2745,
+      "step": 10920
+    },
+    {
+      "epoch": 7.807142857142857,
+      "grad_norm": 2.882291316986084,
+      "learning_rate": 1.1370370370370372e-05,
+      "loss": 0.2353,
+      "step": 10930
+    },
+    {
+      "epoch": 7.814285714285714,
+      "grad_norm": 2.3573696613311768,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 0.2047,
+      "step": 10940
+    },
+    {
+      "epoch": 7.821428571428571,
+      "grad_norm": 1.370251178741455,
+      "learning_rate": 1.1296296296296297e-05,
+      "loss": 0.1855,
+      "step": 10950
+    },
+    {
+      "epoch": 7.828571428571428,
+      "grad_norm": 1.1445660591125488,
+      "learning_rate": 1.125925925925926e-05,
+      "loss": 0.2316,
+      "step": 10960
+    },
+    {
+      "epoch": 7.835714285714285,
+      "grad_norm": 2.094175100326538,
+      "learning_rate": 1.1222222222222224e-05,
+      "loss": 0.477,
+      "step": 10970
+    },
+    {
+      "epoch": 7.8428571428571425,
+      "grad_norm": 2.1148130893707275,
+      "learning_rate": 1.1185185185185187e-05,
+      "loss": 0.2581,
+      "step": 10980
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 1.89934241771698,
+      "learning_rate": 1.1148148148148148e-05,
+      "loss": 0.2908,
+      "step": 10990
+    },
+    {
+      "epoch": 7.857142857142857,
+      "grad_norm": 1.6785616874694824,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.2222,
+      "step": 11000
+    },
+    {
+      "epoch": 7.857142857142857,
+      "eval_loss": 0.34114697575569153,
+      "eval_rouge1": 0.91,
+      "eval_rouge2": 0.8517,
+      "eval_rougeL": 0.9073,
+      "eval_runtime": 122.2351,
+      "eval_samples_per_second": 11.453,
+      "eval_steps_per_second": 5.727,
+      "step": 11000
+    },
+    {
+      "epoch": 7.864285714285714,
+      "grad_norm": 2.335857629776001,
+      "learning_rate": 1.1074074074074075e-05,
+      "loss": 0.262,
+      "step": 11010
+    },
+    {
+      "epoch": 7.871428571428572,
+      "grad_norm": 1.9699100255966187,
+      "learning_rate": 1.1037037037037038e-05,
+      "loss": 0.2056,
+      "step": 11020
+    },
+    {
+      "epoch": 7.878571428571428,
+      "grad_norm": 0.8576107025146484,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.1929,
+      "step": 11030
+    },
+    {
+      "epoch": 7.885714285714286,
+      "grad_norm": 0.7365075945854187,
+      "learning_rate": 1.0962962962962963e-05,
+      "loss": 0.2494,
+      "step": 11040
+    },
+    {
+      "epoch": 7.892857142857143,
+      "grad_norm": 2.7551143169403076,
+      "learning_rate": 1.0925925925925926e-05,
+      "loss": 0.3239,
+      "step": 11050
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 1.1980743408203125,
+      "learning_rate": 1.088888888888889e-05,
+      "loss": 0.1836,
+      "step": 11060
+    },
+    {
+      "epoch": 7.9071428571428575,
+      "grad_norm": 1.525614619255066,
+      "learning_rate": 1.0851851851851853e-05,
+      "loss": 0.3089,
+      "step": 11070
+    },
+    {
+      "epoch": 7.914285714285715,
+      "grad_norm": 2.0727596282958984,
+      "learning_rate": 1.0814814814814814e-05,
+      "loss": 0.1867,
+      "step": 11080
+    },
+    {
+      "epoch": 7.921428571428572,
+      "grad_norm": 1.242550015449524,
+      "learning_rate": 1.0777777777777778e-05,
+      "loss": 0.222,
+      "step": 11090
+    },
+    {
+      "epoch": 7.928571428571429,
+      "grad_norm": 1.3312640190124512,
+      "learning_rate": 1.074074074074074e-05,
+      "loss": 0.3334,
+      "step": 11100
+    },
+    {
+      "epoch": 7.935714285714286,
+      "grad_norm": 1.4483474493026733,
+      "learning_rate": 1.0703703703703704e-05,
+      "loss": 0.2966,
+      "step": 11110
+    },
+    {
+      "epoch": 7.942857142857143,
+      "grad_norm": 1.5403432846069336,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.2591,
+      "step": 11120
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 1.6620547771453857,
+      "learning_rate": 1.0629629629629629e-05,
+      "loss": 0.2216,
+      "step": 11130
+    },
+    {
+      "epoch": 7.957142857142857,
+      "grad_norm": 1.2060527801513672,
+      "learning_rate": 1.0592592592592592e-05,
+      "loss": 0.2403,
+      "step": 11140
+    },
+    {
+      "epoch": 7.964285714285714,
+      "grad_norm": 1.4476624727249146,
+      "learning_rate": 1.0555555555555555e-05,
+      "loss": 0.2739,
+      "step": 11150
+    },
+    {
+      "epoch": 7.9714285714285715,
+      "grad_norm": 2.2425661087036133,
+      "learning_rate": 1.0518518518518519e-05,
+      "loss": 0.2981,
+      "step": 11160
+    },
+    {
+      "epoch": 7.978571428571429,
+      "grad_norm": 2.2100632190704346,
+      "learning_rate": 1.0481481481481482e-05,
+      "loss": 0.2178,
+      "step": 11170
+    },
+    {
+      "epoch": 7.985714285714286,
+      "grad_norm": 1.080759882926941,
+      "learning_rate": 1.0444444444444445e-05,
+      "loss": 0.2947,
+      "step": 11180
+    },
+    {
+      "epoch": 7.992857142857143,
+      "grad_norm": 2.9972569942474365,
+      "learning_rate": 1.0407407407407407e-05,
+      "loss": 0.203,
+      "step": 11190
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.7628710269927979,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.308,
+      "step": 11200
+    },
+    {
+      "epoch": 8.007142857142858,
+      "grad_norm": 1.569732666015625,
+      "learning_rate": 1.0333333333333333e-05,
+      "loss": 0.2003,
+      "step": 11210
+    },
+    {
+      "epoch": 8.014285714285714,
+      "grad_norm": 0.8212767839431763,
+      "learning_rate": 1.0296296296296296e-05,
+      "loss": 0.1957,
+      "step": 11220
+    },
+    {
+      "epoch": 8.021428571428572,
+      "grad_norm": 1.7055811882019043,
+      "learning_rate": 1.025925925925926e-05,
+      "loss": 0.2274,
+      "step": 11230
+    },
+    {
+      "epoch": 8.028571428571428,
+      "grad_norm": 1.6140356063842773,
+      "learning_rate": 1.0222222222222223e-05,
+      "loss": 0.2182,
+      "step": 11240
+    },
+    {
+      "epoch": 8.035714285714286,
+      "grad_norm": 1.6154979467391968,
+      "learning_rate": 1.0185185185185185e-05,
+      "loss": 0.1217,
+      "step": 11250
+    },
+    {
+      "epoch": 8.042857142857143,
+      "grad_norm": 2.4283053874969482,
+      "learning_rate": 1.0148148148148148e-05,
+      "loss": 0.2687,
+      "step": 11260
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 2.201909065246582,
+      "learning_rate": 1.0111111111111111e-05,
+      "loss": 0.2836,
+      "step": 11270
+    },
+    {
+      "epoch": 8.057142857142857,
+      "grad_norm": 1.902273178100586,
+      "learning_rate": 1.0074074074074074e-05,
+      "loss": 0.2076,
+      "step": 11280
+    },
+    {
+      "epoch": 8.064285714285715,
+      "grad_norm": 1.7851389646530151,
+      "learning_rate": 1.0037037037037038e-05,
+      "loss": 0.2045,
+      "step": 11290
+    },
+    {
+      "epoch": 8.071428571428571,
+      "grad_norm": 1.1988000869750977,
+      "learning_rate": 1e-05,
+      "loss": 0.1832,
+      "step": 11300
+    },
+    {
+      "epoch": 8.07857142857143,
+      "grad_norm": 0.6530731320381165,
+      "learning_rate": 9.962962962962962e-06,
+      "loss": 0.1788,
+      "step": 11310
+    },
+    {
+      "epoch": 8.085714285714285,
+      "grad_norm": 1.5712918043136597,
+      "learning_rate": 9.925925925925926e-06,
+      "loss": 0.1601,
+      "step": 11320
+    },
+    {
+      "epoch": 8.092857142857143,
+      "grad_norm": 1.3664653301239014,
+      "learning_rate": 9.888888888888889e-06,
+      "loss": 0.2745,
+      "step": 11330
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 1.084404706954956,
+      "learning_rate": 9.851851851851852e-06,
+      "loss": 0.2595,
+      "step": 11340
+    },
+    {
+      "epoch": 8.107142857142858,
+      "grad_norm": 1.9523823261260986,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 0.2105,
+      "step": 11350
+    },
+    {
+      "epoch": 8.114285714285714,
+      "grad_norm": 1.2386913299560547,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 0.1199,
+      "step": 11360
+    },
+    {
+      "epoch": 8.121428571428572,
+      "grad_norm": 1.6026146411895752,
+      "learning_rate": 9.74074074074074e-06,
+      "loss": 0.2419,
+      "step": 11370
+    },
+    {
+      "epoch": 8.128571428571428,
+      "grad_norm": 1.3624472618103027,
+      "learning_rate": 9.703703703703703e-06,
+      "loss": 0.2094,
+      "step": 11380
+    },
+    {
+      "epoch": 8.135714285714286,
+      "grad_norm": 1.7777026891708374,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 0.2033,
+      "step": 11390
+    },
+    {
+      "epoch": 8.142857142857142,
+      "grad_norm": 1.5583858489990234,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.2229,
+      "step": 11400
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.7537804841995239,
+      "learning_rate": 9.592592592592593e-06,
+      "loss": 0.2201,
+      "step": 11410
+    },
+    {
+      "epoch": 8.157142857142857,
+      "grad_norm": 1.3313623666763306,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 0.1802,
+      "step": 11420
+    },
+    {
+      "epoch": 8.164285714285715,
+      "grad_norm": 2.136382818222046,
+      "learning_rate": 9.51851851851852e-06,
+      "loss": 0.1686,
+      "step": 11430
+    },
+    {
+      "epoch": 8.17142857142857,
+      "grad_norm": 1.4156885147094727,
+      "learning_rate": 9.481481481481481e-06,
+      "loss": 0.2291,
+      "step": 11440
+    },
+    {
+      "epoch": 8.178571428571429,
+      "grad_norm": 1.5296056270599365,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.2476,
+      "step": 11450
+    },
+    {
+      "epoch": 8.185714285714285,
+      "grad_norm": 1.5632902383804321,
+      "learning_rate": 9.407407407407408e-06,
+      "loss": 0.2304,
+      "step": 11460
+    },
+    {
+      "epoch": 8.192857142857143,
+      "grad_norm": 0.9542272686958313,
+      "learning_rate": 9.370370370370371e-06,
+      "loss": 0.1693,
+      "step": 11470
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 3.356255292892456,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.3996,
+      "step": 11480
+    },
+    {
+      "epoch": 8.207142857142857,
+      "grad_norm": 1.6759045124053955,
+      "learning_rate": 9.296296296296298e-06,
+      "loss": 0.2134,
+      "step": 11490
+    },
+    {
+      "epoch": 8.214285714285714,
+      "grad_norm": 1.2791472673416138,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.248,
+      "step": 11500
+    },
+    {
+      "epoch": 8.221428571428572,
+      "grad_norm": 1.07367742061615,
+      "learning_rate": 9.222222222222222e-06,
+      "loss": 0.1731,
+      "step": 11510
+    },
+    {
+      "epoch": 8.228571428571428,
+      "grad_norm": 0.5693773627281189,
+      "learning_rate": 9.185185185185186e-06,
+      "loss": 0.2758,
+      "step": 11520
+    },
+    {
+      "epoch": 8.235714285714286,
+      "grad_norm": 1.7820035219192505,
+      "learning_rate": 9.148148148148149e-06,
+      "loss": 0.2147,
+      "step": 11530
+    },
+    {
+      "epoch": 8.242857142857142,
+      "grad_norm": 1.6593891382217407,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 0.2481,
+      "step": 11540
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 1.4112298488616943,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 0.3307,
+      "step": 11550
+    },
+    {
+      "epoch": 8.257142857142856,
+      "grad_norm": 1.0225239992141724,
+      "learning_rate": 9.037037037037037e-06,
+      "loss": 0.1092,
+      "step": 11560
+    },
+    {
+      "epoch": 8.264285714285714,
+      "grad_norm": 1.4713934659957886,
+      "learning_rate": 9e-06,
+      "loss": 0.2332,
+      "step": 11570
+    },
+    {
+      "epoch": 8.271428571428572,
+      "grad_norm": 1.6922743320465088,
+      "learning_rate": 8.962962962962963e-06,
+      "loss": 0.2106,
+      "step": 11580
+    },
+    {
+      "epoch": 8.278571428571428,
+      "grad_norm": 2.4201695919036865,
+      "learning_rate": 8.925925925925927e-06,
+      "loss": 0.1746,
+      "step": 11590
+    },
+    {
+      "epoch": 8.285714285714286,
+      "grad_norm": 2.3649351596832275,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 0.2204,
+      "step": 11600
+    },
+    {
+      "epoch": 8.292857142857143,
+      "grad_norm": 1.0867241621017456,
+      "learning_rate": 8.851851851851853e-06,
+      "loss": 0.2246,
+      "step": 11610
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 1.1103533506393433,
+      "learning_rate": 8.814814814814815e-06,
+      "loss": 0.2952,
+      "step": 11620
+    },
+    {
+      "epoch": 8.307142857142857,
+      "grad_norm": 1.9086233377456665,
+      "learning_rate": 8.777777777777778e-06,
+      "loss": 0.1627,
+      "step": 11630
+    },
+    {
+      "epoch": 8.314285714285715,
+      "grad_norm": 1.5733546018600464,
+      "learning_rate": 8.740740740740741e-06,
+      "loss": 0.2493,
+      "step": 11640
+    },
+    {
+      "epoch": 8.321428571428571,
+      "grad_norm": 1.514758586883545,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 0.1708,
+      "step": 11650
+    },
+    {
+      "epoch": 8.32857142857143,
+      "grad_norm": 4.691562175750732,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.2622,
+      "step": 11660
+    },
+    {
+      "epoch": 8.335714285714285,
+      "grad_norm": 1.0987350940704346,
+      "learning_rate": 8.629629629629631e-06,
+      "loss": 0.1291,
+      "step": 11670
+    },
+    {
+      "epoch": 8.342857142857143,
+      "grad_norm": 1.4016727209091187,
+      "learning_rate": 8.592592592592593e-06,
+      "loss": 0.2075,
+      "step": 11680
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.69717937707901,
+      "learning_rate": 8.555555555555556e-06,
+      "loss": 0.1609,
+      "step": 11690
+    },
+    {
+      "epoch": 8.357142857142858,
+      "grad_norm": 2.023461103439331,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.235,
+      "step": 11700
+    },
+    {
+      "epoch": 8.364285714285714,
+      "grad_norm": 2.0078303813934326,
+      "learning_rate": 8.481481481481482e-06,
+      "loss": 0.2081,
+      "step": 11710
+    },
+    {
+      "epoch": 8.371428571428572,
+      "grad_norm": 1.4724724292755127,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 0.2148,
+      "step": 11720
+    },
+    {
+      "epoch": 8.378571428571428,
+      "grad_norm": 2.9564125537872314,
+      "learning_rate": 8.407407407407409e-06,
+      "loss": 0.2295,
+      "step": 11730
+    },
+    {
+      "epoch": 8.385714285714286,
+      "grad_norm": 1.4059520959854126,
+      "learning_rate": 8.37037037037037e-06,
+      "loss": 0.1398,
+      "step": 11740
+    },
+    {
+      "epoch": 8.392857142857142,
+      "grad_norm": 2.894953966140747,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2699,
+      "step": 11750
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 2.5017454624176025,
+      "learning_rate": 8.296296296296297e-06,
+      "loss": 0.2322,
+      "step": 11760
+    },
+    {
+      "epoch": 8.407142857142857,
+      "grad_norm": 0.9503372311592102,
+      "learning_rate": 8.25925925925926e-06,
+      "loss": 0.1356,
+      "step": 11770
+    },
+    {
+      "epoch": 8.414285714285715,
+      "grad_norm": 1.63711416721344,
+      "learning_rate": 8.222222222222223e-06,
+      "loss": 0.1929,
+      "step": 11780
+    },
+    {
+      "epoch": 8.42142857142857,
+      "grad_norm": 0.5972274541854858,
+      "learning_rate": 8.185185185185187e-06,
+      "loss": 0.2533,
+      "step": 11790
+    },
+    {
+      "epoch": 8.428571428571429,
+      "grad_norm": 1.0893709659576416,
+      "learning_rate": 8.14814814814815e-06,
+      "loss": 0.2089,
+      "step": 11800
+    },
+    {
+      "epoch": 8.435714285714285,
+      "grad_norm": 1.5523369312286377,
+      "learning_rate": 8.111111111111112e-06,
+      "loss": 0.1589,
+      "step": 11810
+    },
+    {
+      "epoch": 8.442857142857143,
+      "grad_norm": 1.5510472059249878,
+      "learning_rate": 8.074074074074075e-06,
+      "loss": 0.2162,
+      "step": 11820
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 2.1272058486938477,
+      "learning_rate": 8.037037037037038e-06,
+      "loss": 0.2172,
+      "step": 11830
+    },
+    {
+      "epoch": 8.457142857142857,
+      "grad_norm": 2.2862300872802734,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.2341,
+      "step": 11840
+    },
+    {
+      "epoch": 8.464285714285714,
+      "grad_norm": 1.126405954360962,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 0.1474,
+      "step": 11850
+    },
+    {
+      "epoch": 8.471428571428572,
+      "grad_norm": 2.000462293624878,
+      "learning_rate": 7.925925925925926e-06,
+      "loss": 0.2948,
+      "step": 11860
+    },
+    {
+      "epoch": 8.478571428571428,
+      "grad_norm": 1.4239530563354492,
+      "learning_rate": 7.88888888888889e-06,
+      "loss": 0.2267,
+      "step": 11870
+    },
+    {
+      "epoch": 8.485714285714286,
+      "grad_norm": 2.479355812072754,
+      "learning_rate": 7.851851851851853e-06,
+      "loss": 0.1857,
+      "step": 11880
+    },
+    {
+      "epoch": 8.492857142857144,
+      "grad_norm": 1.9741036891937256,
+      "learning_rate": 7.814814814814816e-06,
+      "loss": 0.2049,
+      "step": 11890
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 1.996099591255188,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 0.2404,
+      "step": 11900
+    },
+    {
+      "epoch": 8.507142857142856,
+      "grad_norm": 1.185811161994934,
+      "learning_rate": 7.74074074074074e-06,
+      "loss": 0.1476,
+      "step": 11910
+    },
+    {
+      "epoch": 8.514285714285714,
+      "grad_norm": 1.7766746282577515,
+      "learning_rate": 7.703703703703704e-06,
+      "loss": 0.1483,
+      "step": 11920
+    },
+    {
+      "epoch": 8.521428571428572,
+      "grad_norm": 1.0609338283538818,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 0.3414,
+      "step": 11930
+    },
+    {
+      "epoch": 8.528571428571428,
+      "grad_norm": 1.352981448173523,
+      "learning_rate": 7.629629629629629e-06,
+      "loss": 0.2453,
+      "step": 11940
+    },
+    {
+      "epoch": 8.535714285714286,
+      "grad_norm": 2.01450252532959,
+      "learning_rate": 7.592592592592593e-06,
+      "loss": 0.0983,
+      "step": 11950
+    },
+    {
+      "epoch": 8.542857142857143,
+      "grad_norm": 0.9895955920219421,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 0.2821,
+      "step": 11960
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 2.1095712184906006,
+      "learning_rate": 7.518518518518519e-06,
+      "loss": 0.2147,
+      "step": 11970
+    },
+    {
+      "epoch": 8.557142857142857,
+      "grad_norm": 1.3148187398910522,
+      "learning_rate": 7.481481481481483e-06,
+      "loss": 0.1834,
+      "step": 11980
+    },
+    {
+      "epoch": 8.564285714285715,
+      "grad_norm": 2.1209769248962402,
+      "learning_rate": 7.444444444444444e-06,
+      "loss": 0.2369,
+      "step": 11990
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 2.559124708175659,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.3642,
+      "step": 12000
+    },
+    {
+      "epoch": 8.571428571428571,
+      "eval_loss": 0.3329981863498688,
+      "eval_rouge1": 0.9114,
+      "eval_rouge2": 0.8538,
+      "eval_rougeL": 0.9087,
+      "eval_runtime": 122.2883,
+      "eval_samples_per_second": 11.448,
+      "eval_steps_per_second": 5.724,
+      "step": 12000
+    },
+    {
+      "epoch": 8.57857142857143,
+      "grad_norm": 2.582113027572632,
+      "learning_rate": 7.370370370370371e-06,
+      "loss": 0.3589,
+      "step": 12010
+    },
+    {
+      "epoch": 8.585714285714285,
+      "grad_norm": 1.1606924533843994,
+      "learning_rate": 7.333333333333334e-06,
+      "loss": 0.2036,
+      "step": 12020
+    },
+    {
+      "epoch": 8.592857142857143,
+      "grad_norm": 1.990123987197876,
+      "learning_rate": 7.296296296296297e-06,
+      "loss": 0.2347,
+      "step": 12030
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 1.3593547344207764,
+      "learning_rate": 7.2592592592592605e-06,
+      "loss": 0.1475,
+      "step": 12040
+    },
+    {
+      "epoch": 8.607142857142858,
+      "grad_norm": 0.7821537256240845,
+      "learning_rate": 7.222222222222222e-06,
+      "loss": 0.2889,
+      "step": 12050
+    },
+    {
+      "epoch": 8.614285714285714,
+      "grad_norm": 2.2297816276550293,
+      "learning_rate": 7.185185185185185e-06,
+      "loss": 0.2495,
+      "step": 12060
+    },
+    {
+      "epoch": 8.621428571428572,
+      "grad_norm": 1.9694111347198486,
+      "learning_rate": 7.1481481481481486e-06,
+      "loss": 0.2182,
+      "step": 12070
+    },
+    {
+      "epoch": 8.628571428571428,
+      "grad_norm": 1.3611793518066406,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 0.2116,
+      "step": 12080
+    },
+    {
+      "epoch": 8.635714285714286,
+      "grad_norm": 0.9108260869979858,
+      "learning_rate": 7.074074074074074e-06,
+      "loss": 0.2229,
+      "step": 12090
+    },
+    {
+      "epoch": 8.642857142857142,
+      "grad_norm": 2.577470302581787,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.3132,
+      "step": 12100
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.837846040725708,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.1791,
+      "step": 12110
+    },
+    {
+      "epoch": 8.657142857142857,
+      "grad_norm": 2.7253611087799072,
+      "learning_rate": 6.962962962962963e-06,
+      "loss": 0.1998,
+      "step": 12120
+    },
+    {
+      "epoch": 8.664285714285715,
+      "grad_norm": 1.6284639835357666,
+      "learning_rate": 6.925925925925926e-06,
+      "loss": 0.2917,
+      "step": 12130
+    },
+    {
+      "epoch": 8.67142857142857,
+      "grad_norm": 1.0358765125274658,
+      "learning_rate": 6.888888888888889e-06,
+      "loss": 0.2949,
+      "step": 12140
+    },
+    {
+      "epoch": 8.678571428571429,
+      "grad_norm": 1.9100443124771118,
+      "learning_rate": 6.851851851851852e-06,
+      "loss": 0.1639,
+      "step": 12150
+    },
+    {
+      "epoch": 8.685714285714285,
+      "grad_norm": 1.4746047258377075,
+      "learning_rate": 6.814814814814815e-06,
+      "loss": 0.1798,
+      "step": 12160
+    },
+    {
+      "epoch": 8.692857142857143,
+      "grad_norm": 1.1177834272384644,
+      "learning_rate": 6.777777777777779e-06,
+      "loss": 0.2687,
+      "step": 12170
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.8141186833381653,
+      "learning_rate": 6.74074074074074e-06,
+      "loss": 0.1627,
+      "step": 12180
+    },
+    {
+      "epoch": 8.707142857142857,
+      "grad_norm": 1.4017144441604614,
+      "learning_rate": 6.703703703703703e-06,
+      "loss": 0.1938,
+      "step": 12190
+    },
+    {
+      "epoch": 8.714285714285714,
+      "grad_norm": 0.7198919653892517,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.222,
+      "step": 12200
+    },
+    {
+      "epoch": 8.721428571428572,
+      "grad_norm": 1.4436593055725098,
+      "learning_rate": 6.62962962962963e-06,
+      "loss": 0.2821,
+      "step": 12210
+    },
+    {
+      "epoch": 8.728571428571428,
+      "grad_norm": 1.8222748041152954,
+      "learning_rate": 6.592592592592593e-06,
+      "loss": 0.3652,
+      "step": 12220
+    },
+    {
+      "epoch": 8.735714285714286,
+      "grad_norm": 1.033396601676941,
+      "learning_rate": 6.555555555555556e-06,
+      "loss": 0.1775,
+      "step": 12230
+    },
+    {
+      "epoch": 8.742857142857144,
+      "grad_norm": 1.3893368244171143,
+      "learning_rate": 6.518518518518518e-06,
+      "loss": 0.3292,
+      "step": 12240
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.7294946908950806,
+      "learning_rate": 6.481481481481481e-06,
+      "loss": 0.3136,
+      "step": 12250
+    },
+    {
+      "epoch": 8.757142857142856,
+      "grad_norm": 1.5931612253189087,
+      "learning_rate": 6.4444444444444445e-06,
+      "loss": 0.2498,
+      "step": 12260
+    },
+    {
+      "epoch": 8.764285714285714,
+      "grad_norm": 1.6048915386199951,
+      "learning_rate": 6.407407407407408e-06,
+      "loss": 0.2996,
+      "step": 12270
+    },
+    {
+      "epoch": 8.771428571428572,
+      "grad_norm": 0.9885932803153992,
+      "learning_rate": 6.370370370370371e-06,
+      "loss": 0.1647,
+      "step": 12280
+    },
+    {
+      "epoch": 8.778571428571428,
+      "grad_norm": 4.228821277618408,
+      "learning_rate": 6.333333333333334e-06,
+      "loss": 0.2533,
+      "step": 12290
+    },
+    {
+      "epoch": 8.785714285714286,
+      "grad_norm": 2.150721788406372,
+      "learning_rate": 6.296296296296296e-06,
+      "loss": 0.2922,
+      "step": 12300
+    },
+    {
+      "epoch": 8.792857142857143,
+      "grad_norm": 2.0247929096221924,
+      "learning_rate": 6.259259259259259e-06,
+      "loss": 0.1613,
+      "step": 12310
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.5003380179405212,
+      "learning_rate": 6.222222222222222e-06,
+      "loss": 0.1459,
+      "step": 12320
+    },
+    {
+      "epoch": 8.807142857142857,
+      "grad_norm": 0.9374644756317139,
+      "learning_rate": 6.1851851851851856e-06,
+      "loss": 0.2384,
+      "step": 12330
+    },
+    {
+      "epoch": 8.814285714285715,
+      "grad_norm": 1.5784926414489746,
+      "learning_rate": 6.148148148148149e-06,
+      "loss": 0.2458,
+      "step": 12340
+    },
+    {
+      "epoch": 8.821428571428571,
+      "grad_norm": 1.6696372032165527,
+      "learning_rate": 6.111111111111111e-06,
+      "loss": 0.2787,
+      "step": 12350
+    },
+    {
+      "epoch": 8.82857142857143,
+      "grad_norm": 1.2524793148040771,
+      "learning_rate": 6.0740740740740745e-06,
+      "loss": 0.2445,
+      "step": 12360
+    },
+    {
+      "epoch": 8.835714285714285,
+      "grad_norm": 2.176713466644287,
+      "learning_rate": 6.037037037037038e-06,
+      "loss": 0.1896,
+      "step": 12370
+    },
+    {
+      "epoch": 8.842857142857143,
+      "grad_norm": 1.964752197265625,
+      "learning_rate": 6e-06,
+      "loss": 0.261,
+      "step": 12380
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 2.006638526916504,
+      "learning_rate": 5.962962962962963e-06,
+      "loss": 0.2415,
+      "step": 12390
+    },
+    {
+      "epoch": 8.857142857142858,
+      "grad_norm": 1.9313900470733643,
+      "learning_rate": 5.925925925925927e-06,
+      "loss": 0.3772,
+      "step": 12400
+    },
+    {
+      "epoch": 8.864285714285714,
+      "grad_norm": 2.112170934677124,
+      "learning_rate": 5.888888888888889e-06,
+      "loss": 0.2396,
+      "step": 12410
+    },
+    {
+      "epoch": 8.871428571428572,
+      "grad_norm": 2.7346787452697754,
+      "learning_rate": 5.851851851851852e-06,
+      "loss": 0.325,
+      "step": 12420
+    },
+    {
+      "epoch": 8.878571428571428,
+      "grad_norm": 3.1768178939819336,
+      "learning_rate": 5.814814814814816e-06,
+      "loss": 0.2393,
+      "step": 12430
+    },
+    {
+      "epoch": 8.885714285714286,
+      "grad_norm": 2.164926767349243,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 0.2444,
+      "step": 12440
+    },
+    {
+      "epoch": 8.892857142857142,
+      "grad_norm": 0.8274826407432556,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 0.2662,
+      "step": 12450
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 1.024776816368103,
+      "learning_rate": 5.7037037037037045e-06,
+      "loss": 0.2059,
+      "step": 12460
+    },
+    {
+      "epoch": 8.907142857142857,
+      "grad_norm": 1.1089740991592407,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 0.2529,
+      "step": 12470
+    },
+    {
+      "epoch": 8.914285714285715,
+      "grad_norm": 1.3669365644454956,
+      "learning_rate": 5.62962962962963e-06,
+      "loss": 0.2432,
+      "step": 12480
+    },
+    {
+      "epoch": 8.92142857142857,
+      "grad_norm": 1.9711703062057495,
+      "learning_rate": 5.592592592592593e-06,
+      "loss": 0.127,
+      "step": 12490
+    },
+    {
+      "epoch": 8.928571428571429,
+      "grad_norm": 1.4796557426452637,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.2244,
+      "step": 12500
+    },
+    {
+      "epoch": 8.935714285714285,
+      "grad_norm": 2.029303789138794,
+      "learning_rate": 5.518518518518519e-06,
+      "loss": 0.2939,
+      "step": 12510
+    },
+    {
+      "epoch": 8.942857142857143,
+      "grad_norm": 2.0809414386749268,
+      "learning_rate": 5.4814814814814815e-06,
+      "loss": 0.2146,
+      "step": 12520
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 1.2555458545684814,
+      "learning_rate": 5.444444444444445e-06,
+      "loss": 0.2116,
+      "step": 12530
+    },
+    {
+      "epoch": 8.957142857142857,
+      "grad_norm": 2.3402106761932373,
+      "learning_rate": 5.407407407407407e-06,
+      "loss": 0.1587,
+      "step": 12540
+    },
+    {
+      "epoch": 8.964285714285714,
+      "grad_norm": 1.2514257431030273,
+      "learning_rate": 5.37037037037037e-06,
+      "loss": 0.1598,
+      "step": 12550
+    },
+    {
+      "epoch": 8.971428571428572,
+      "grad_norm": 3.1680376529693604,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.2925,
+      "step": 12560
+    },
+    {
+      "epoch": 8.978571428571428,
+      "grad_norm": 2.482211112976074,
+      "learning_rate": 5.296296296296296e-06,
+      "loss": 0.2152,
+      "step": 12570
+    },
+    {
+      "epoch": 8.985714285714286,
+      "grad_norm": 2.3472371101379395,
+      "learning_rate": 5.259259259259259e-06,
+      "loss": 0.2079,
+      "step": 12580
+    },
+    {
+      "epoch": 8.992857142857144,
+      "grad_norm": 2.52826189994812,
+      "learning_rate": 5.2222222222222226e-06,
+      "loss": 0.2124,
+      "step": 12590
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.8535226583480835,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.232,
+      "step": 12600
+    },
+    {
+      "epoch": 9.007142857142858,
+      "grad_norm": 2.8780601024627686,
+      "learning_rate": 5.148148148148148e-06,
+      "loss": 0.3111,
+      "step": 12610
+    },
+    {
+      "epoch": 9.014285714285714,
+      "grad_norm": 1.5909632444381714,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 0.2788,
+      "step": 12620
+    },
+    {
+      "epoch": 9.021428571428572,
+      "grad_norm": 1.7076550722122192,
+      "learning_rate": 5.074074074074074e-06,
+      "loss": 0.2575,
+      "step": 12630
+    },
+    {
+      "epoch": 9.028571428571428,
+      "grad_norm": 2.8672637939453125,
+      "learning_rate": 5.037037037037037e-06,
+      "loss": 0.2228,
+      "step": 12640
+    },
+    {
+      "epoch": 9.035714285714286,
+      "grad_norm": 1.618055820465088,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 12650
+    },
+    {
+      "epoch": 9.042857142857143,
+      "grad_norm": 1.407122015953064,
+      "learning_rate": 4.962962962962963e-06,
+      "loss": 0.2321,
+      "step": 12660
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 1.6875501871109009,
+      "learning_rate": 4.925925925925926e-06,
+      "loss": 0.2938,
+      "step": 12670
+    },
+    {
+      "epoch": 9.057142857142857,
+      "grad_norm": 1.0872751474380493,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 0.181,
+      "step": 12680
+    },
+    {
+      "epoch": 9.064285714285715,
+      "grad_norm": 1.689308524131775,
+      "learning_rate": 4.851851851851852e-06,
+      "loss": 0.1147,
+      "step": 12690
+    },
+    {
+      "epoch": 9.071428571428571,
+      "grad_norm": 2.5880138874053955,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.2757,
+      "step": 12700
+    },
+    {
+      "epoch": 9.07857142857143,
+      "grad_norm": 2.9612958431243896,
+      "learning_rate": 4.777777777777778e-06,
+      "loss": 0.2847,
+      "step": 12710
+    },
+    {
+      "epoch": 9.085714285714285,
+      "grad_norm": 1.7059327363967896,
+      "learning_rate": 4.740740740740741e-06,
+      "loss": 0.195,
+      "step": 12720
+    },
+    {
+      "epoch": 9.092857142857143,
+      "grad_norm": 1.702331304550171,
+      "learning_rate": 4.703703703703704e-06,
+      "loss": 0.2408,
+      "step": 12730
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 1.814587950706482,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.3085,
+      "step": 12740
+    },
+    {
+      "epoch": 9.107142857142858,
+      "grad_norm": 2.863785743713379,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 0.2667,
+      "step": 12750
+    },
+    {
+      "epoch": 9.114285714285714,
+      "grad_norm": 1.6600861549377441,
+      "learning_rate": 4.592592592592593e-06,
+      "loss": 0.1376,
+      "step": 12760
+    },
+    {
+      "epoch": 9.121428571428572,
+      "grad_norm": 1.3329426050186157,
+      "learning_rate": 4.555555555555556e-06,
+      "loss": 0.2547,
+      "step": 12770
+    },
+    {
+      "epoch": 9.128571428571428,
+      "grad_norm": 4.032371997833252,
+      "learning_rate": 4.5185185185185185e-06,
+      "loss": 0.4649,
+      "step": 12780
+    },
+    {
+      "epoch": 9.135714285714286,
+      "grad_norm": 2.709066390991211,
+      "learning_rate": 4.481481481481482e-06,
+      "loss": 0.2336,
+      "step": 12790
+    },
+    {
+      "epoch": 9.142857142857142,
+      "grad_norm": 1.063931941986084,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 0.1731,
+      "step": 12800
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 1.3110073804855347,
+      "learning_rate": 4.407407407407407e-06,
+      "loss": 0.1381,
+      "step": 12810
+    },
+    {
+      "epoch": 9.157142857142857,
+      "grad_norm": 1.9574129581451416,
+      "learning_rate": 4.370370370370371e-06,
+      "loss": 0.1958,
+      "step": 12820
+    },
+    {
+      "epoch": 9.164285714285715,
+      "grad_norm": 2.5129504203796387,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 0.2162,
+      "step": 12830
+    },
+    {
+      "epoch": 9.17142857142857,
+      "grad_norm": 1.842850923538208,
+      "learning_rate": 4.296296296296296e-06,
+      "loss": 0.2746,
+      "step": 12840
+    },
+    {
+      "epoch": 9.178571428571429,
+      "grad_norm": 1.1451313495635986,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 0.2442,
+      "step": 12850
+    },
+    {
+      "epoch": 9.185714285714285,
+      "grad_norm": 2.0765175819396973,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 0.1531,
+      "step": 12860
+    },
+    {
+      "epoch": 9.192857142857143,
+      "grad_norm": 0.9250321984291077,
+      "learning_rate": 4.185185185185185e-06,
+      "loss": 0.2876,
+      "step": 12870
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 1.5332380533218384,
+      "learning_rate": 4.1481481481481485e-06,
+      "loss": 0.2255,
+      "step": 12880
+    },
+    {
+      "epoch": 9.207142857142857,
+      "grad_norm": 3.1344316005706787,
+      "learning_rate": 4.111111111111112e-06,
+      "loss": 0.2654,
+      "step": 12890
+    },
+    {
+      "epoch": 9.214285714285714,
+      "grad_norm": 1.4092166423797607,
+      "learning_rate": 4.074074074074075e-06,
+      "loss": 0.2358,
+      "step": 12900
+    },
+    {
+      "epoch": 9.221428571428572,
+      "grad_norm": 1.517716646194458,
+      "learning_rate": 4.037037037037037e-06,
+      "loss": 0.2455,
+      "step": 12910
+    },
+    {
+      "epoch": 9.228571428571428,
+      "grad_norm": 0.7355996966362,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.1761,
+      "step": 12920
+    },
+    {
+      "epoch": 9.235714285714286,
+      "grad_norm": 2.077259063720703,
+      "learning_rate": 3.962962962962963e-06,
+      "loss": 0.2791,
+      "step": 12930
+    },
+    {
+      "epoch": 9.242857142857142,
+      "grad_norm": 1.3175309896469116,
+      "learning_rate": 3.925925925925926e-06,
+      "loss": 0.2296,
+      "step": 12940
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 1.0608943700790405,
+      "learning_rate": 3.888888888888889e-06,
+      "loss": 0.1609,
+      "step": 12950
+    },
+    {
+      "epoch": 9.257142857142856,
+      "grad_norm": 2.2581288814544678,
+      "learning_rate": 3.851851851851852e-06,
+      "loss": 0.2192,
+      "step": 12960
+    },
+    {
+      "epoch": 9.264285714285714,
+      "grad_norm": 1.672400951385498,
+      "learning_rate": 3.814814814814814e-06,
+      "loss": 0.1548,
+      "step": 12970
+    },
+    {
+      "epoch": 9.271428571428572,
+      "grad_norm": 0.7743004560470581,
+      "learning_rate": 3.777777777777778e-06,
+      "loss": 0.3168,
+      "step": 12980
+    },
+    {
+      "epoch": 9.278571428571428,
+      "grad_norm": 0.9588621854782104,
+      "learning_rate": 3.7407407407407413e-06,
+      "loss": 0.1585,
+      "step": 12990
+    },
+    {
+      "epoch": 9.285714285714286,
+      "grad_norm": 1.7508875131607056,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.2137,
+      "step": 13000
+    },
+    {
+      "epoch": 9.285714285714286,
+      "eval_loss": 0.3308734893798828,
+      "eval_rouge1": 0.9118,
+      "eval_rouge2": 0.8545,
+      "eval_rougeL": 0.909,
+      "eval_runtime": 122.198,
+      "eval_samples_per_second": 11.457,
+      "eval_steps_per_second": 5.728,
+      "step": 13000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 14000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 2,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.24700797698048e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}