diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,27100 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999741314639005,
+  "eval_steps": 500,
+  "global_step": 19328,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0346611484738748e-07,
+      "loss": 2.0672,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.173305742369374e-07,
+      "loss": 2.0719,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0346611484738748e-06,
+      "loss": 2.0752,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5519917227108122e-06,
+      "loss": 2.0881,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.0693222969477496e-06,
+      "loss": 2.0211,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.586652871184687e-06,
+      "loss": 2.0231,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.1039834454216244e-06,
+      "loss": 2.0532,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.6213140196585623e-06,
+      "loss": 2.0431,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.138644593895499e-06,
+      "loss": 2.0622,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5,
+      "learning_rate": 4.655975168132437e-06,
+      "loss": 2.0141,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.173305742369374e-06,
+      "loss": 2.0169,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.6906363166063115e-06,
+      "loss": 2.0446,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.207966890843249e-06,
+      "loss": 2.0514,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.725297465080186e-06,
+      "loss": 2.028,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.2426280393171246e-06,
+      "loss": 1.9917,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.75995861355406e-06,
+      "loss": 2.0271,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.277289187790999e-06,
+      "loss": 1.9452,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.794619762027937e-06,
+      "loss": 1.9469,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.311950336264873e-06,
+      "loss": 1.9631,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.82928091050181e-06,
+      "loss": 1.939,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0346611484738748e-05,
+      "loss": 1.939,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.0863942058975686e-05,
+      "loss": 1.9396,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.1381272633212623e-05,
+      "loss": 1.9,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1898603207449561e-05,
+      "loss": 1.9245,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.2415933781686498e-05,
+      "loss": 1.8493,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2933264355923436e-05,
+      "loss": 1.8828,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3450594930160373e-05,
+      "loss": 1.8809,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.3967925504397311e-05,
+      "loss": 1.8241,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.4485256078634249e-05,
+      "loss": 1.8636,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5002586652871187e-05,
+      "loss": 1.8901,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.551991722710812e-05,
+      "loss": 1.8464,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.603724780134506e-05,
+      "loss": 1.7929,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6554578375581997e-05,
+      "loss": 1.8162,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.7071908949818935e-05,
+      "loss": 1.8365,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7589239524055874e-05,
+      "loss": 1.8309,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8106570098292812e-05,
+      "loss": 1.8047,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8623900672529747e-05,
+      "loss": 1.8204,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9141231246766685e-05,
+      "loss": 1.8022,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.965856182100362e-05,
+      "loss": 1.8294,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.0175892395240558e-05,
+      "loss": 1.8189,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.0693222969477496e-05,
+      "loss": 1.8118,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.1210553543714435e-05,
+      "loss": 1.8137,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1727884117951373e-05,
+      "loss": 1.7953,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.224521469218831e-05,
+      "loss": 1.8048,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.2762545266425246e-05,
+      "loss": 1.7902,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.3279875840662184e-05,
+      "loss": 1.8015,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.625,
+      "learning_rate": 2.3797206414899122e-05,
+      "loss": 1.8083,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.4314536989136057e-05,
+      "loss": 1.7645,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.4831867563372996e-05,
+      "loss": 1.7412,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.5349198137609937e-05,
+      "loss": 1.7416,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.5866528711846872e-05,
+      "loss": 1.7907,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.6383859286083807e-05,
+      "loss": 1.7905,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.6901189860320745e-05,
+      "loss": 1.7633,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.7418520434557683e-05,
+      "loss": 1.7678,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.7935851008794622e-05,
+      "loss": 1.7868,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.8453181583031557e-05,
+      "loss": 1.7642,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.8970512157268498e-05,
+      "loss": 1.7907,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.9487842731505433e-05,
+      "loss": 1.7673,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.0005173305742375e-05,
+      "loss": 1.7521,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.052250387997931e-05,
+      "loss": 1.7589,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.103983445421624e-05,
+      "loss": 1.7534,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.1557165028453186e-05,
+      "loss": 1.7469,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.207449560269012e-05,
+      "loss": 1.7285,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.259182617692706e-05,
+      "loss": 1.7777,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.3109156751163994e-05,
+      "loss": 1.6814,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.362648732540093e-05,
+      "loss": 1.739,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.414381789963787e-05,
+      "loss": 1.7609,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.46611484738748e-05,
+      "loss": 1.7366,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.75,
+      "learning_rate": 3.517847904811175e-05,
+      "loss": 1.7704,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.569580962234868e-05,
+      "loss": 1.6942,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.6213140196585624e-05,
+      "loss": 1.6884,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.6730470770822555e-05,
+      "loss": 1.7542,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.724780134505949e-05,
+      "loss": 1.7435,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.776513191929643e-05,
+      "loss": 1.7158,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.828246249353337e-05,
+      "loss": 1.7178,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.879979306777031e-05,
+      "loss": 1.7613,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.931712364200724e-05,
+      "loss": 1.7211,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 380.0,
+      "learning_rate": 3.9834454216244185e-05,
+      "loss": 1.796,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.0351784790481116e-05,
+      "loss": 1.7092,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.086911536471806e-05,
+      "loss": 1.7148,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.138644593895499e-05,
+      "loss": 1.6965,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.190377651319193e-05,
+      "loss": 1.7083,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.242110708742887e-05,
+      "loss": 1.6792,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.293843766166581e-05,
+      "loss": 1.7314,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.3455768235902746e-05,
+      "loss": 1.741,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.397309881013968e-05,
+      "loss": 1.7095,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.449042938437662e-05,
+      "loss": 1.7073,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.5007759958613554e-05,
+      "loss": 1.7033,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.552509053285049e-05,
+      "loss": 1.6986,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.604242110708743e-05,
+      "loss": 1.6743,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.655975168132437e-05,
+      "loss": 1.6705,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.707708225556131e-05,
+      "loss": 1.6937,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.7594412829798245e-05,
+      "loss": 1.7355,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.811174340403518e-05,
+      "loss": 1.6826,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.8629073978272115e-05,
+      "loss": 1.6627,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.914640455250906e-05,
+      "loss": 1.6788,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.966373512674599e-05,
+      "loss": 1.7126,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.018106570098293e-05,
+      "loss": 1.7152,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.0698396275219874e-05,
+      "loss": 1.6685,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.12157268494568e-05,
+      "loss": 1.6562,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.1733057423693744e-05,
+      "loss": 1.651,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.225038799793068e-05,
+      "loss": 1.6997,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.2767718572167614e-05,
+      "loss": 1.6904,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.328504914640455e-05,
+      "loss": 1.6754,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.380237972064149e-05,
+      "loss": 1.7106,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.4319710294878435e-05,
+      "loss": 1.6913,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.483704086911537e-05,
+      "loss": 1.6779,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.5354371443352305e-05,
+      "loss": 1.6658,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.5871702017589243e-05,
+      "loss": 1.6931,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.6389032591826175e-05,
+      "loss": 1.6653,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.75,
+      "learning_rate": 5.690636316606311e-05,
+      "loss": 1.6921,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.742369374030006e-05,
+      "loss": 1.6621,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.7941024314536996e-05,
+      "loss": 1.6868,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.75,
+      "learning_rate": 5.845835488877393e-05,
+      "loss": 1.672,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.75,
+      "learning_rate": 5.8975685463010866e-05,
+      "loss": 1.6741,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.9493016037247804e-05,
+      "loss": 1.6989,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.001034661148475e-05,
+      "loss": 1.6492,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.0527677185721674e-05,
+      "loss": 1.7075,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.104500775995862e-05,
+      "loss": 1.6889,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.156233833419556e-05,
+      "loss": 1.6749,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.207966890843248e-05,
+      "loss": 1.6923,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.259699948266943e-05,
+      "loss": 1.6846,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.311433005690637e-05,
+      "loss": 1.6878,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.363166063114331e-05,
+      "loss": 1.665,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.414899120538024e-05,
+      "loss": 1.6745,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.466632177961717e-05,
+      "loss": 1.6813,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.518365235385413e-05,
+      "loss": 1.6718,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.570098292809105e-05,
+      "loss": 1.6575,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.621831350232799e-05,
+      "loss": 1.6663,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.673564407656493e-05,
+      "loss": 1.659,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.725297465080186e-05,
+      "loss": 1.6705,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.77703052250388e-05,
+      "loss": 1.6884,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.828763579927574e-05,
+      "loss": 1.6974,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.880496637351268e-05,
+      "loss": 1.6686,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.93222969477496e-05,
+      "loss": 1.6708,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.983962752198656e-05,
+      "loss": 1.6826,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.03569580962235e-05,
+      "loss": 1.6771,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.087428867046043e-05,
+      "loss": 1.657,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.139161924469736e-05,
+      "loss": 1.6522,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.19089498189343e-05,
+      "loss": 1.6453,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.242628039317125e-05,
+      "loss": 1.6442,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.294361096740819e-05,
+      "loss": 1.631,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.346094154164511e-05,
+      "loss": 1.6205,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.397827211588205e-05,
+      "loss": 1.6374,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.449560269011899e-05,
+      "loss": 1.6693,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.501293326435593e-05,
+      "loss": 1.6841,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.553026383859286e-05,
+      "loss": 1.6304,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.60475944128298e-05,
+      "loss": 1.6373,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.656492498706674e-05,
+      "loss": 1.6395,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.708225556130368e-05,
+      "loss": 1.6679,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.759958613554062e-05,
+      "loss": 1.6473,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.811691670977755e-05,
+      "loss": 1.6029,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.863424728401448e-05,
+      "loss": 1.6523,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.915157785825143e-05,
+      "loss": 1.6281,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.966890843248837e-05,
+      "loss": 1.6032,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.018623900672531e-05,
+      "loss": 1.6495,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.070356958096223e-05,
+      "loss": 1.6309,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.122090015519917e-05,
+      "loss": 1.6321,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.173823072943612e-05,
+      "loss": 1.6487,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.225556130367305e-05,
+      "loss": 1.648,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.277289187790999e-05,
+      "loss": 1.6622,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.329022245214692e-05,
+      "loss": 1.6737,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.380755302638386e-05,
+      "loss": 1.6437,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.43248836006208e-05,
+      "loss": 1.6797,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.484221417485774e-05,
+      "loss": 1.6351,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.535954474909468e-05,
+      "loss": 1.6527,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.587687532333161e-05,
+      "loss": 1.6279,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.639420589756855e-05,
+      "loss": 1.6513,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.691153647180549e-05,
+      "loss": 1.6415,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.742886704604243e-05,
+      "loss": 1.6349,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.794619762027935e-05,
+      "loss": 1.6161,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.846352819451629e-05,
+      "loss": 1.6249,
+      "step": 855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.898085876875324e-05,
+      "loss": 1.6353,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.949818934299018e-05,
+      "loss": 1.6625,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.001551991722711e-05,
+      "loss": 1.6277,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.053285049146405e-05,
+      "loss": 1.675,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.105018106570098e-05,
+      "loss": 1.654,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.156751163993792e-05,
+      "loss": 1.6671,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.208484221417486e-05,
+      "loss": 1.6186,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.625,
+      "learning_rate": 9.26021727884118e-05,
+      "loss": 1.643,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.625,
+      "learning_rate": 9.311950336264874e-05,
+      "loss": 1.6187,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.363683393688568e-05,
+      "loss": 1.6145,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.415416451112261e-05,
+      "loss": 1.6412,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.467149508535955e-05,
+      "loss": 1.6214,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.518882565959649e-05,
+      "loss": 1.6311,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.570615623383343e-05,
+      "loss": 1.6373,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.622348680807037e-05,
+      "loss": 1.6238,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.625,
+      "learning_rate": 9.67408173823073e-05,
+      "loss": 1.6447,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.725814795654423e-05,
+      "loss": 1.6108,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.777547853078117e-05,
+      "loss": 1.6195,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.829280910501812e-05,
+      "loss": 1.6216,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.881013967925506e-05,
+      "loss": 1.6312,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.932747025349198e-05,
+      "loss": 1.6351,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.984480082772892e-05,
+      "loss": 1.6219,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010036213140196586,
+      "loss": 1.6085,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001008794619762028,
+      "loss": 1.5999,
+      "step": 975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010139679255043975,
+      "loss": 1.638,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010191412312467667,
+      "loss": 1.6279,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001024314536989136,
+      "loss": 1.6012,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010294878427315055,
+      "loss": 1.5978,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010346611484738749,
+      "loss": 1.6574,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010398344542162441,
+      "loss": 1.6144,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010450077599586136,
+      "loss": 1.5865,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010501810657009829,
+      "loss": 1.5854,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010553543714433523,
+      "loss": 1.5843,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 28.125,
+      "learning_rate": 0.00010605276771857218,
+      "loss": 1.737,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001065700982928091,
+      "loss": 1.6076,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010708742886704606,
+      "loss": 1.6308,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010760475944128298,
+      "loss": 1.5846,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010812209001551992,
+      "loss": 1.6332,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010863942058975687,
+      "loss": 1.6076,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001091567511639938,
+      "loss": 1.6014,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010967408173823073,
+      "loss": 1.6394,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011019141231246769,
+      "loss": 1.5853,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00011070874288670461,
+      "loss": 1.5903,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011122607346094154,
+      "loss": 1.6,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011174340403517849,
+      "loss": 1.636,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011226073460941543,
+      "loss": 1.632,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011277806518365235,
+      "loss": 1.5872,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001132953957578893,
+      "loss": 1.6298,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011381272633212623,
+      "loss": 1.6073,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011433005690636318,
+      "loss": 1.5932,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011484738748060012,
+      "loss": 1.5957,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011536471805483704,
+      "loss": 1.5781,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011588204862907399,
+      "loss": 1.5761,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011639937920331092,
+      "loss": 1.5804,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011691670977754786,
+      "loss": 1.6005,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011743404035178481,
+      "loss": 1.5513,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011795137092602173,
+      "loss": 1.6244,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011846870150025866,
+      "loss": 1.5941,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011898603207449561,
+      "loss": 1.6306,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011950336264873255,
+      "loss": 1.6149,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001200206932229695,
+      "loss": 1.603,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012053802379720642,
+      "loss": 1.5644,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012105535437144335,
+      "loss": 1.5883,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001215726849456803,
+      "loss": 1.5846,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012209001551991724,
+      "loss": 1.6083,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012260734609415415,
+      "loss": 1.6051,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012312467666839111,
+      "loss": 1.6224,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012364200724262805,
+      "loss": 1.6303,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012415933781686496,
+      "loss": 1.5796,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012467666839110193,
+      "loss": 1.5645,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012519399896533887,
+      "loss": 1.591,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012571132953957578,
+      "loss": 1.6042,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012622866011381274,
+      "loss": 1.5599,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012674599068804966,
+      "loss": 1.6363,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012726332126228662,
+      "loss": 1.5934,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012778065183652356,
+      "loss": 1.5754,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012829798241076047,
+      "loss": 1.586,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012881531298499744,
+      "loss": 1.5841,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012933264355923435,
+      "loss": 1.5874,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012984997413347129,
+      "loss": 1.592,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013036730470770825,
+      "loss": 1.6185,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013088463528194516,
+      "loss": 1.603,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001314019658561821,
+      "loss": 1.542,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013191929643041904,
+      "loss": 1.5686,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013243662700465598,
+      "loss": 1.5893,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013295395757889294,
+      "loss": 1.6039,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013347128815312985,
+      "loss": 1.6043,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001339886187273668,
+      "loss": 1.5886,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013450594930160373,
+      "loss": 1.6044,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013502327987584067,
+      "loss": 1.5665,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001355406104500776,
+      "loss": 1.5608,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013605794102431454,
+      "loss": 1.6247,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013657527159855148,
+      "loss": 1.5377,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013709260217278842,
+      "loss": 1.5695,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013760993274702536,
+      "loss": 1.5705,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001381272633212623,
+      "loss": 1.5698,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001386445938954992,
+      "loss": 1.5721,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013916192446973617,
+      "loss": 1.5982,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001396792550439731,
+      "loss": 1.5812,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014019658561821005,
+      "loss": 1.62,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000140713916192447,
+      "loss": 1.6233,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001412312467666839,
+      "loss": 1.5828,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014174857734092087,
+      "loss": 1.594,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001422659079151578,
+      "loss": 1.6036,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014278323848939471,
+      "loss": 1.616,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014330056906363168,
+      "loss": 1.5825,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001438178996378686,
+      "loss": 1.5786,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014433523021210553,
+      "loss": 1.5681,
+      "step": 1395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001448525607863425,
+      "loss": 1.5683,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001453698913605794,
+      "loss": 1.5987,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014588722193481637,
+      "loss": 1.5762,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00014640455250905328,
+      "loss": 1.5804,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014692188308329022,
+      "loss": 1.5967,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014743921365752719,
+      "loss": 1.564,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001479565442317641,
+      "loss": 1.5913,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014847387480600104,
+      "loss": 1.5789,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014899120538023797,
+      "loss": 1.5692,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001495085359544749,
+      "loss": 1.5682,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015002586652871185,
+      "loss": 1.5949,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001505431971029488,
+      "loss": 1.6124,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015106052767718573,
+      "loss": 1.611,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015157785825142266,
+      "loss": 1.5875,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001520951888256596,
+      "loss": 1.5513,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015261251939989654,
+      "loss": 1.5774,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015312984997413348,
+      "loss": 1.58,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015364718054837042,
+      "loss": 1.584,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015416451112260736,
+      "loss": 1.6072,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001546818416968443,
+      "loss": 1.545,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015519917227108123,
+      "loss": 1.59,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015571650284531817,
+      "loss": 1.5463,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001562338334195551,
+      "loss": 1.5767,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015675116399379205,
+      "loss": 1.6077,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015726849456802896,
+      "loss": 1.5653,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015778582514226592,
+      "loss": 1.5685,
+      "step": 1525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015830315571650286,
+      "loss": 1.5647,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001588204862907398,
+      "loss": 1.5662,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015933781686497674,
+      "loss": 1.5859,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015985514743921365,
+      "loss": 1.5762,
+      "step": 1545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016037247801345062,
+      "loss": 1.5691,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016088980858768755,
+      "loss": 1.5798,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016140713916192446,
+      "loss": 1.5793,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016192446973616143,
+      "loss": 1.5979,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016244180031039834,
+      "loss": 1.5795,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016295913088463528,
+      "loss": 1.5537,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016347646145887224,
+      "loss": 1.5306,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016399379203310916,
+      "loss": 1.5812,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001645111226073461,
+      "loss": 1.5892,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016502845318158303,
+      "loss": 1.5867,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016554578375581997,
+      "loss": 1.5385,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016606311433005694,
+      "loss": 1.5428,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016658044490429385,
+      "loss": 1.5596,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016709777547853079,
+      "loss": 1.5558,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016761510605276772,
+      "loss": 1.5684,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016813243662700466,
+      "loss": 1.572,
+      "step": 1625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001686497672012416,
+      "loss": 1.5505,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016916709777547854,
+      "loss": 1.5732,
+      "step": 1635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016968442834971548,
+      "loss": 1.5629,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017020175892395241,
+      "loss": 1.5986,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017071908949818935,
+      "loss": 1.5384,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001712364200724263,
+      "loss": 1.5552,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017175375064666323,
+      "loss": 1.5812,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017227108122090017,
+      "loss": 1.5385,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001727884117951371,
+      "loss": 1.5383,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017330574236937404,
+      "loss": 1.5549,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017382307294361098,
+      "loss": 1.5509,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001743404035178479,
+      "loss": 1.6073,
+      "step": 1685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017485773409208486,
+      "loss": 1.5538,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001753750646663218,
+      "loss": 1.595,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001758923952405587,
+      "loss": 1.5658,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017640972581479567,
+      "loss": 1.5508,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017692705638903259,
+      "loss": 1.5406,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017744438696326955,
+      "loss": 1.5676,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001779617175375065,
+      "loss": 1.5622,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001784790481117434,
+      "loss": 1.5609,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017899637868598037,
+      "loss": 1.578,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017951370926021728,
+      "loss": 1.5461,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018003103983445421,
+      "loss": 1.5818,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018054837040869118,
+      "loss": 1.5628,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.3125,
+      "learning_rate": 0.0001810657009829281,
+      "loss": 1.568,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018158303155716503,
+      "loss": 1.5412,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018210036213140197,
+      "loss": 1.6057,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001826176927056389,
+      "loss": 1.5805,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018313502327987584,
+      "loss": 1.5736,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018365235385411278,
+      "loss": 1.5586,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00018416968442834972,
+      "loss": 1.5524,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018468701500258666,
+      "loss": 1.5908,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001852043455768236,
+      "loss": 1.5742,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018572167615106054,
+      "loss": 1.5759,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00018623900672529747,
+      "loss": 1.6109,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001867563372995344,
+      "loss": 1.545,
+      "step": 1805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018727366787377135,
+      "loss": 1.5727,
+      "step": 1810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001877909984480083,
+      "loss": 1.5265,
+      "step": 1815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018830832902224523,
+      "loss": 1.5541,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018882565959648216,
+      "loss": 1.5675,
+      "step": 1825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001893429901707191,
+      "loss": 1.5542,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018986032074495604,
+      "loss": 1.5244,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019037765131919298,
+      "loss": 1.5473,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019089498189342992,
+      "loss": 1.5554,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019141231246766686,
+      "loss": 1.5603,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001919296430419038,
+      "loss": 1.5157,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019244697361614073,
+      "loss": 1.5392,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019296430419037764,
+      "loss": 1.5601,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001934816347646146,
+      "loss": 1.5473,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019399896533885155,
+      "loss": 1.5782,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019451629591308846,
+      "loss": 1.5659,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019503362648732542,
+      "loss": 1.5565,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019555095706156234,
+      "loss": 1.5245,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019606828763579927,
+      "loss": 1.5682,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019658561821003624,
+      "loss": 1.5439,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.0,
+      "learning_rate": 0.00019710294878427315,
+      "loss": 1.5547,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019762027935851012,
+      "loss": 1.5742,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019813760993274703,
+      "loss": 1.524,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019865494050698396,
+      "loss": 1.5328,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019917227108122093,
+      "loss": 1.5554,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019968960165545784,
+      "loss": 1.57,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.5,
+      "learning_rate": 0.00019999999347649694,
+      "loss": 1.5967,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019999992008709735,
+      "loss": 1.5606,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019999976515397937,
+      "loss": 1.5559,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019999952867726936,
+      "loss": 1.5675,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001999992106571601,
+      "loss": 1.5612,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019999881109391098,
+      "loss": 1.5634,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001999983299878478,
+      "loss": 1.5404,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019999776733936286,
+      "loss": 1.5875,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019999712314891496,
+      "loss": 1.5651,
+      "step": 1975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019999639741702943,
+      "loss": 1.5764,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019999559014429802,
+      "loss": 1.5411,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019999470133137906,
+      "loss": 1.5388,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019999373097899728,
+      "loss": 1.5659,
+      "step": 1995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019999267908794394,
+      "loss": 1.5484,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019999154565907682,
+      "loss": 1.5623,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019999033069332013,
+      "loss": 1.5429,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001999890341916646,
+      "loss": 1.5245,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001999876561551675,
+      "loss": 1.5415,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019998619658495245,
+      "loss": 1.5505,
+      "step": 2025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019998465548220972,
+      "loss": 1.5406,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019998303284819594,
+      "loss": 1.5452,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019998132868423427,
+      "loss": 1.5519,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019997954299171434,
+      "loss": 1.5671,
+      "step": 2045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001999776757720923,
+      "loss": 1.5425,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019997572702689073,
+      "loss": 1.5056,
+      "step": 2055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019997369675769873,
+      "loss": 1.511,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019997158496617184,
+      "loss": 1.5355,
+      "step": 2065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019996939165403208,
+      "loss": 1.5132,
+      "step": 2070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000199967116823068,
+      "loss": 1.5543,
+      "step": 2075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019996476047513454,
+      "loss": 1.5596,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001999623226121532,
+      "loss": 1.5632,
+      "step": 2085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001999598032361119,
+      "loss": 1.5241,
+      "step": 2090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019995720234906498,
+      "loss": 1.5281,
+      "step": 2095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019995451995313335,
+      "loss": 1.5327,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019995175605050434,
+      "loss": 1.5475,
+      "step": 2105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001999489106434317,
+      "loss": 1.5329,
+      "step": 2110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001999459837342357,
+      "loss": 1.5358,
+      "step": 2115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019994297532530312,
+      "loss": 1.5547,
+      "step": 2120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019993988541908703,
+      "loss": 1.552,
+      "step": 2125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019993671401810712,
+      "loss": 1.561,
+      "step": 2130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019993346112494946,
+      "loss": 1.5502,
+      "step": 2135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019993012674226655,
+      "loss": 1.5832,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001999267108727774,
+      "loss": 1.5401,
+      "step": 2145
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019992321351926744,
+      "loss": 1.5634,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019991963468458853,
+      "loss": 1.5423,
+      "step": 2155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019991597437165899,
+      "loss": 1.5138,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019991223258346362,
+      "loss": 1.4988,
+      "step": 2165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019990840932305353,
+      "loss": 1.5531,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001999045045935464,
+      "loss": 1.522,
+      "step": 2175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019990051839812633,
+      "loss": 1.5382,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019989645074004376,
+      "loss": 1.5066,
+      "step": 2185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001998923016226156,
+      "loss": 1.5335,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001998880710492253,
+      "loss": 1.5735,
+      "step": 2195
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001998837590233225,
+      "loss": 1.5011,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019987936554842346,
+      "loss": 1.5417,
+      "step": 2205
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019987489062811076,
+      "loss": 1.4801,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019987033426603344,
+      "loss": 1.5865,
+      "step": 2215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00019986569646590692,
+      "loss": 1.5352,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019986097723151305,
+      "loss": 1.5486,
+      "step": 2225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019985617656670005,
+      "loss": 1.5046,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019985129447538258,
+      "loss": 1.5207,
+      "step": 2235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019984633096154167,
+      "loss": 1.5484,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019984128602922477,
+      "loss": 1.4857,
+      "step": 2245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019983615968254573,
+      "loss": 1.5185,
+      "step": 2250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001998309519256847,
+      "loss": 1.5341,
+      "step": 2255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019982566276288834,
+      "loss": 1.5127,
+      "step": 2260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019982029219846962,
+      "loss": 1.5261,
+      "step": 2265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019981484023680787,
+      "loss": 1.5209,
+      "step": 2270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019980930688234886,
+      "loss": 1.5647,
+      "step": 2275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019980369213960472,
+      "loss": 1.5466,
+      "step": 2280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019979799601315388,
+      "loss": 1.5414,
+      "step": 2285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019979221850764117,
+      "loss": 1.5479,
+      "step": 2290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001997863596277778,
+      "loss": 1.5523,
+      "step": 2295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019978041937834137,
+      "loss": 1.5381,
+      "step": 2300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001997743977641757,
+      "loss": 1.5089,
+      "step": 2305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.5625,
+      "learning_rate": 0.00019976829479019113,
+      "loss": 1.5425,
+      "step": 2310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001997621104613642,
+      "loss": 1.5251,
+      "step": 2315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019975584478273782,
+      "loss": 1.5214,
+      "step": 2320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019974949775942134,
+      "loss": 1.5123,
+      "step": 2325
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019974306939659026,
+      "loss": 1.5431,
+      "step": 2330
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019973655969948663,
+      "loss": 1.5478,
+      "step": 2335
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019972996867341863,
+      "loss": 1.5273,
+      "step": 2340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019972329632376084,
+      "loss": 1.5277,
+      "step": 2345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019971654265595415,
+      "loss": 1.5046,
+      "step": 2350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019970970767550577,
+      "loss": 1.5588,
+      "step": 2355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001997027913879892,
+      "loss": 1.6231,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001996957937990442,
+      "loss": 1.5189,
+      "step": 2365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019968871491437691,
+      "loss": 1.5559,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019968155473975974,
+      "loss": 1.5194,
+      "step": 2375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001996743132810313,
+      "loss": 1.5081,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001996669905440966,
+      "loss": 1.5502,
+      "step": 2385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001996595865349269,
+      "loss": 1.5206,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019965210125955966,
+      "loss": 1.5631,
+      "step": 2395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019964453472409867,
+      "loss": 1.4963,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019963688693471396,
+      "loss": 1.5388,
+      "step": 2405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019962915789764182,
+      "loss": 1.5142,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019962134761918488,
+      "loss": 1.5213,
+      "step": 2415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019961345610571183,
+      "loss": 1.5471,
+      "step": 2420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019960548336365774,
+      "loss": 1.507,
+      "step": 2425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019959742939952392,
+      "loss": 1.5371,
+      "step": 2430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019958929421987783,
+      "loss": 1.5465,
+      "step": 2435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019958107783135326,
+      "loss": 1.4912,
+      "step": 2440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019957278024065013,
+      "loss": 1.5464,
+      "step": 2445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019956440145453458,
+      "loss": 1.5109,
+      "step": 2450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019955594147983905,
+      "loss": 1.5218,
+      "step": 2455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019954740032346208,
+      "loss": 1.4954,
+      "step": 2460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001995387779923685,
+      "loss": 1.5223,
+      "step": 2465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001995300744935892,
+      "loss": 1.5373,
+      "step": 2470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019952128983422146,
+      "loss": 1.5302,
+      "step": 2475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019951242402142848,
+      "loss": 1.5762,
+      "step": 2480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001995034770624399,
+      "loss": 1.4838,
+      "step": 2485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019949444896455137,
+      "loss": 1.5001,
+      "step": 2490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019948533973512472,
+      "loss": 1.5255,
+      "step": 2495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000199476149381588,
+      "loss": 1.5226,
+      "step": 2500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001994668779114353,
+      "loss": 1.5017,
+      "step": 2505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019945752533222704,
+      "loss": 1.5007,
+      "step": 2510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019944809165158955,
+      "loss": 1.5411,
+      "step": 2515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001994385768772155,
+      "loss": 1.5656,
+      "step": 2520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019942898101686356,
+      "loss": 1.5273,
+      "step": 2525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019941930407835857,
+      "loss": 1.5197,
+      "step": 2530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019940954606959143,
+      "loss": 1.4927,
+      "step": 2535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019939970699851925,
+      "loss": 1.5054,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001993897868731651,
+      "loss": 1.5288,
+      "step": 2545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019937978570161834,
+      "loss": 1.4935,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019936970349203423,
+      "loss": 1.5252,
+      "step": 2555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019935954025263416,
+      "loss": 1.5234,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019934929599170568,
+      "loss": 1.5277,
+      "step": 2565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019933897071760235,
+      "loss": 1.5002,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019932856443874374,
+      "loss": 1.4713,
+      "step": 2575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019931807716361554,
+      "loss": 1.5328,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019930750890076947,
+      "loss": 1.5025,
+      "step": 2585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001992968596588233,
+      "loss": 1.5213,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019928612944646084,
+      "loss": 1.4832,
+      "step": 2595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019927531827243188,
+      "loss": 1.5389,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001992644261455523,
+      "loss": 1.5366,
+      "step": 2605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001992534530747039,
+      "loss": 1.5488,
+      "step": 2610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019924239906883457,
+      "loss": 1.4925,
+      "step": 2615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019923126413695817,
+      "loss": 1.4764,
+      "step": 2620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019922004828815454,
+      "loss": 1.5102,
+      "step": 2625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001992087515315695,
+      "loss": 1.5488,
+      "step": 2630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019919737387641485,
+      "loss": 1.5403,
+      "step": 2635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019918591533196834,
+      "loss": 1.5077,
+      "step": 2640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019917437590757375,
+      "loss": 1.4922,
+      "step": 2645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019916275561264075,
+      "loss": 1.5049,
+      "step": 2650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019915105445664493,
+      "loss": 1.5126,
+      "step": 2655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019913927244912788,
+      "loss": 1.5371,
+      "step": 2660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001991274095996971,
+      "loss": 1.4777,
+      "step": 2665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019911546591802604,
+      "loss": 1.5001,
+      "step": 2670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019910344141385396,
+      "loss": 1.5312,
+      "step": 2675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019909133609698616,
+      "loss": 1.4799,
+      "step": 2680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019907914997729372,
+      "loss": 1.4944,
+      "step": 2685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019906688306471366,
+      "loss": 1.4979,
+      "step": 2690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019905453536924893,
+      "loss": 1.5127,
+      "step": 2695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001990421069009683,
+      "loss": 1.5091,
+      "step": 2700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001990295976700064,
+      "loss": 1.5043,
+      "step": 2705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019901700768656372,
+      "loss": 1.5113,
+      "step": 2710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001990043369609066,
+      "loss": 1.5135,
+      "step": 2715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019899158550336729,
+      "loss": 1.5443,
+      "step": 2720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019897875332434376,
+      "loss": 1.4857,
+      "step": 2725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019896584043429988,
+      "loss": 1.5084,
+      "step": 2730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019895284684376524,
+      "loss": 1.4913,
+      "step": 2735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001989397725633354,
+      "loss": 1.5016,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019892661760367156,
+      "loss": 1.5035,
+      "step": 2745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019891338197550081,
+      "loss": 1.5148,
+      "step": 2750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019890006568961597,
+      "loss": 1.4888,
+      "step": 2755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019888666875687565,
+      "loss": 1.5469,
+      "step": 2760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019887319118820418,
+      "loss": 1.5647,
+      "step": 2765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001988596329945917,
+      "loss": 1.5234,
+      "step": 2770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001988459941870941,
+      "loss": 1.5043,
+      "step": 2775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019883227477683296,
+      "loss": 1.5245,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019881847477499557,
+      "loss": 1.5148,
+      "step": 2785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019880459419283503,
+      "loss": 1.5096,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001987906330416701,
+      "loss": 1.5181,
+      "step": 2795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019877659133288515,
+      "loss": 1.5406,
+      "step": 2800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001987624690779304,
+      "loss": 1.5078,
+      "step": 2805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019874826628832164,
+      "loss": 1.5313,
+      "step": 2810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019873398297564037,
+      "loss": 1.4807,
+      "step": 2815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001987196191515337,
+      "loss": 1.5308,
+      "step": 2820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001987051748277145,
+      "loss": 1.5101,
+      "step": 2825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019869065001596118,
+      "loss": 1.4953,
+      "step": 2830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019867604472811786,
+      "loss": 1.4817,
+      "step": 2835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019866135897609423,
+      "loss": 1.4956,
+      "step": 2840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019864659277186555,
+      "loss": 1.5151,
+      "step": 2845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001986317461274728,
+      "loss": 1.4996,
+      "step": 2850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019861681905502246,
+      "loss": 1.5088,
+      "step": 2855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001986018115666867,
+      "loss": 1.507,
+      "step": 2860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019858672367470312,
+      "loss": 1.516,
+      "step": 2865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.000198571555391375,
+      "loss": 1.5231,
+      "step": 2870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019855630672907108,
+      "loss": 1.4969,
+      "step": 2875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019854097770022577,
+      "loss": 1.4924,
+      "step": 2880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001985255683173389,
+      "loss": 1.5237,
+      "step": 2885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019851007859297585,
+      "loss": 1.5364,
+      "step": 2890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019849450853976755,
+      "loss": 1.5116,
+      "step": 2895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001984788581704104,
+      "loss": 1.4999,
+      "step": 2900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001984631274976663,
+      "loss": 1.4907,
+      "step": 2905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019844731653436264,
+      "loss": 1.5079,
+      "step": 2910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001984314252933923,
+      "loss": 1.4708,
+      "step": 2915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019841545378771356,
+      "loss": 1.4926,
+      "step": 2920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001983994020303502,
+      "loss": 1.4863,
+      "step": 2925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019838327003439147,
+      "loss": 1.4937,
+      "step": 2930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019836705781299196,
+      "loss": 1.535,
+      "step": 2935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019835076537937178,
+      "loss": 1.4663,
+      "step": 2940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019833439274681634,
+      "loss": 1.5312,
+      "step": 2945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019831793992867652,
+      "loss": 1.5367,
+      "step": 2950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001983014069383686,
+      "loss": 1.5108,
+      "step": 2955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019828479378937417,
+      "loss": 1.5443,
+      "step": 2960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019826810049524026,
+      "loss": 1.506,
+      "step": 2965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019825132706957917,
+      "loss": 1.5365,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019823447352606858,
+      "loss": 1.4662,
+      "step": 2975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019821753987845156,
+      "loss": 1.5271,
+      "step": 2980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001982005261405364,
+      "loss": 1.4906,
+      "step": 2985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001981834323261968,
+      "loss": 1.4914,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019816625844937163,
+      "loss": 1.5608,
+      "step": 2995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001981490045240652,
+      "loss": 1.4996,
+      "step": 3000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019813167056434693,
+      "loss": 1.4952,
+      "step": 3005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019811425658435166,
+      "loss": 1.5259,
+      "step": 3010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019809676259827935,
+      "loss": 1.4858,
+      "step": 3015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001980791886203953,
+      "loss": 1.5333,
+      "step": 3020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019806153466502997,
+      "loss": 1.5455,
+      "step": 3025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019804380074657906,
+      "loss": 1.5162,
+      "step": 3030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019802598687950352,
+      "loss": 1.5035,
+      "step": 3035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019800809307832942,
+      "loss": 1.4887,
+      "step": 3040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019799011935764803,
+      "loss": 1.4994,
+      "step": 3045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001979720657321158,
+      "loss": 1.5093,
+      "step": 3050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019795393221645437,
+      "loss": 1.5287,
+      "step": 3055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019793571882545047,
+      "loss": 1.4979,
+      "step": 3060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019791742557395602,
+      "loss": 1.5138,
+      "step": 3065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000197899052476888,
+      "loss": 1.525,
+      "step": 3070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019788059954922856,
+      "loss": 1.4849,
+      "step": 3075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019786206680602486,
+      "loss": 1.531,
+      "step": 3080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019784345426238927,
+      "loss": 1.528,
+      "step": 3085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019782476193349905,
+      "loss": 1.5096,
+      "step": 3090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019780598983459678,
+      "loss": 1.5177,
+      "step": 3095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019778713798098983,
+      "loss": 1.5012,
+      "step": 3100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019776820638805077,
+      "loss": 1.4774,
+      "step": 3105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001977491950712171,
+      "loss": 1.5024,
+      "step": 3110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001977301040459914,
+      "loss": 1.5284,
+      "step": 3115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019771093332794117,
+      "loss": 1.5153,
+      "step": 3120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000197691682932699,
+      "loss": 1.5114,
+      "step": 3125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019767235287596237,
+      "loss": 1.4915,
+      "step": 3130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001976529431734937,
+      "loss": 1.4751,
+      "step": 3135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019763345384112043,
+      "loss": 1.5189,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001976138848947349,
+      "loss": 1.5114,
+      "step": 3145
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019759423635029434,
+      "loss": 1.5104,
+      "step": 3150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019757450822382094,
+      "loss": 1.5031,
+      "step": 3155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019755470053140178,
+      "loss": 1.4643,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001975348132891888,
+      "loss": 1.483,
+      "step": 3165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019751484651339877,
+      "loss": 1.4759,
+      "step": 3170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019749480022031337,
+      "loss": 1.4971,
+      "step": 3175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019747467442627912,
+      "loss": 1.4768,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019745446914770732,
+      "loss": 1.4869,
+      "step": 3185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019743418440107418,
+      "loss": 1.5524,
+      "step": 3190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019741382020292063,
+      "loss": 1.4935,
+      "step": 3195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019739337656985234,
+      "loss": 1.5014,
+      "step": 3200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001973728535185399,
+      "loss": 1.5191,
+      "step": 3205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019735225106571854,
+      "loss": 1.4759,
+      "step": 3210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019733156922818835,
+      "loss": 1.5358,
+      "step": 3215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019731080802281396,
+      "loss": 1.4926,
+      "step": 3220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019728996746652496,
+      "loss": 1.4928,
+      "step": 3225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019726904757631544,
+      "loss": 1.4744,
+      "step": 3230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001972480483692443,
+      "loss": 1.516,
+      "step": 3235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019722696986243515,
+      "loss": 1.4968,
+      "step": 3240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019720581207307612,
+      "loss": 1.5516,
+      "step": 3245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001971845750184201,
+      "loss": 1.4762,
+      "step": 3250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019716325871578462,
+      "loss": 1.4957,
+      "step": 3255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001971418631825517,
+      "loss": 1.4826,
+      "step": 3260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019712038843616817,
+      "loss": 1.4764,
+      "step": 3265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019709883449414535,
+      "loss": 1.4964,
+      "step": 3270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019707720137405907,
+      "loss": 1.5065,
+      "step": 3275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019705548909354983,
+      "loss": 1.4792,
+      "step": 3280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019703369767032266,
+      "loss": 1.4646,
+      "step": 3285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001970118271221471,
+      "loss": 1.5114,
+      "step": 3290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001969898774668572,
+      "loss": 1.4886,
+      "step": 3295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019696784872235158,
+      "loss": 1.509,
+      "step": 3300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001969457409065933,
+      "loss": 1.4883,
+      "step": 3305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019692355403760987,
+      "loss": 1.5194,
+      "step": 3310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019690128813349333,
+      "loss": 1.5052,
+      "step": 3315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019687894321240016,
+      "loss": 1.4941,
+      "step": 3320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019685651929255123,
+      "loss": 1.5108,
+      "step": 3325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001968340163922319,
+      "loss": 1.5232,
+      "step": 3330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019681143452979178,
+      "loss": 1.5024,
+      "step": 3335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001967887737236451,
+      "loss": 1.4631,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019676603399227023,
+      "loss": 1.5092,
+      "step": 3345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001967432153542101,
+      "loss": 1.5084,
+      "step": 3350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019672031782807178,
+      "loss": 1.4948,
+      "step": 3355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001966973414325269,
+      "loss": 1.4951,
+      "step": 3360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019667428618631126,
+      "loss": 1.5071,
+      "step": 3365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019665115210822489,
+      "loss": 1.4793,
+      "step": 3370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019662793921713226,
+      "loss": 1.5236,
+      "step": 3375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019660464753196207,
+      "loss": 1.4882,
+      "step": 3380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019658127707170716,
+      "loss": 1.4904,
+      "step": 3385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019655782785542476,
+      "loss": 1.4852,
+      "step": 3390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001965342999022362,
+      "loss": 1.4834,
+      "step": 3395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001965106932313271,
+      "loss": 1.5002,
+      "step": 3400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001964870078619472,
+      "loss": 1.4868,
+      "step": 3405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019646324381341045,
+      "loss": 1.514,
+      "step": 3410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000196439401105095,
+      "loss": 1.4797,
+      "step": 3415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019641547975644304,
+      "loss": 1.4759,
+      "step": 3420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019639147978696097,
+      "loss": 1.5039,
+      "step": 3425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001963674012162193,
+      "loss": 1.4789,
+      "step": 3430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019634324406385252,
+      "loss": 1.5134,
+      "step": 3435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019631900834955935,
+      "loss": 1.4714,
+      "step": 3440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019629469409310253,
+      "loss": 1.5027,
+      "step": 3445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019627030131430875,
+      "loss": 1.5045,
+      "step": 3450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001962458300330689,
+      "loss": 1.5231,
+      "step": 3455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001962212802693377,
+      "loss": 1.5213,
+      "step": 3460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000196196652043134,
+      "loss": 1.5295,
+      "step": 3465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001961719453745406,
+      "loss": 1.5081,
+      "step": 3470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001961471602837042,
+      "loss": 1.4992,
+      "step": 3475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019612229679083555,
+      "loss": 1.5028,
+      "step": 3480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001960973549162093,
+      "loss": 1.5116,
+      "step": 3485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019607233468016392,
+      "loss": 1.4824,
+      "step": 3490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019604723610310194,
+      "loss": 1.5318,
+      "step": 3495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019602205920548965,
+      "loss": 1.5108,
+      "step": 3500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001959968040078572,
+      "loss": 1.5155,
+      "step": 3505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019597147053079873,
+      "loss": 1.4939,
+      "step": 3510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019594605879497202,
+      "loss": 1.4775,
+      "step": 3515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019592056882109885,
+      "loss": 1.5044,
+      "step": 3520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019589500062996463,
+      "loss": 1.4672,
+      "step": 3525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019586935424241873,
+      "loss": 1.4761,
+      "step": 3530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019584362967937406,
+      "loss": 1.4901,
+      "step": 3535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019581782696180748,
+      "loss": 1.4989,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001957919461107595,
+      "loss": 1.5152,
+      "step": 3545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019576598714733431,
+      "loss": 1.5188,
+      "step": 3550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019573995009269988,
+      "loss": 1.4681,
+      "step": 3555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019571383496808775,
+      "loss": 1.4673,
+      "step": 3560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019568764179479323,
+      "loss": 1.4654,
+      "step": 3565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001956613705941752,
+      "loss": 1.5073,
+      "step": 3570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019563502138765618,
+      "loss": 1.4786,
+      "step": 3575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019560859419672237,
+      "loss": 1.5432,
+      "step": 3580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019558208904292342,
+      "loss": 1.497,
+      "step": 3585
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001955555059478727,
+      "loss": 1.4615,
+      "step": 3590
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019552884493324703,
+      "loss": 1.4729,
+      "step": 3595
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019550210602078684,
+      "loss": 1.5153,
+      "step": 3600
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000195475289232296,
+      "loss": 1.4976,
+      "step": 3605
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019544839458964202,
+      "loss": 1.5,
+      "step": 3610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001954214221147557,
+      "loss": 1.4885,
+      "step": 3615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019539437182963153,
+      "loss": 1.4852,
+      "step": 3620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019536724375632727,
+      "loss": 1.4751,
+      "step": 3625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019534003791696417,
+      "loss": 1.4512,
+      "step": 3630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019531275433372694,
+      "loss": 1.4634,
+      "step": 3635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019528539302886362,
+      "loss": 1.4936,
+      "step": 3640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019525795402468567,
+      "loss": 1.4721,
+      "step": 3645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019523043734356787,
+      "loss": 1.5107,
+      "step": 3650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019520284300794837,
+      "loss": 1.5082,
+      "step": 3655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019517517104032864,
+      "loss": 1.4966,
+      "step": 3660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019514742146327344,
+      "loss": 1.4804,
+      "step": 3665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019511959429941087,
+      "loss": 1.5073,
+      "step": 3670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001950916895714322,
+      "loss": 1.5222,
+      "step": 3675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000195063707302092,
+      "loss": 1.4928,
+      "step": 3680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001950356475142081,
+      "loss": 1.4753,
+      "step": 3685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019500751023066154,
+      "loss": 1.4971,
+      "step": 3690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019497929547439643,
+      "loss": 1.4678,
+      "step": 3695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001949510032684202,
+      "loss": 1.4907,
+      "step": 3700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019492263363580343,
+      "loss": 1.5243,
+      "step": 3705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019489418659967975,
+      "loss": 1.485,
+      "step": 3710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019486566218324597,
+      "loss": 1.4388,
+      "step": 3715
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019483706040976194,
+      "loss": 1.4807,
+      "step": 3720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001948083813025506,
+      "loss": 1.4988,
+      "step": 3725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001947796248849981,
+      "loss": 1.4606,
+      "step": 3730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001947507911805534,
+      "loss": 1.4861,
+      "step": 3735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019472188021272868,
+      "loss": 1.4952,
+      "step": 3740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019469289200509896,
+      "loss": 1.4767,
+      "step": 3745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019466382658130232,
+      "loss": 1.5096,
+      "step": 3750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019463468396503989,
+      "loss": 1.4809,
+      "step": 3755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001946054641800756,
+      "loss": 1.4876,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019457616725023635,
+      "loss": 1.4583,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000194546793199412,
+      "loss": 1.4932,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019451734205155527,
+      "loss": 1.4892,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019448781383068174,
+      "loss": 1.5205,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001944582085608698,
+      "loss": 1.5026,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019442852626626076,
+      "loss": 1.4821,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001943987669710586,
+      "loss": 1.5057,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001943689306995303,
+      "loss": 1.466,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019433901747600537,
+      "loss": 1.4606,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019430902732487626,
+      "loss": 1.4627,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00019427896027059802,
+      "loss": 1.4864,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019424881633768853,
+      "loss": 1.497,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019421859555072822,
+      "loss": 1.4953,
+      "step": 3825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001941882979343603,
+      "loss": 1.4895,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019415792351329058,
+      "loss": 1.4654,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019412747231228753,
+      "loss": 1.5066,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019409694435618222,
+      "loss": 1.5092,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019406633966986828,
+      "loss": 1.4891,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001940356582783019,
+      "loss": 1.4741,
+      "step": 3855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001940049002065019,
+      "loss": 1.4906,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019397406547954954,
+      "loss": 1.4775,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019394315412258868,
+      "loss": 1.4981,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019391216616082552,
+      "loss": 1.4714,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001938811016195289,
+      "loss": 1.504,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019384996052402995,
+      "loss": 1.5086,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019381874289972238,
+      "loss": 1.4597,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001937874487720621,
+      "loss": 1.527,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019375607816656768,
+      "loss": 1.4881,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001937246311088198,
+      "loss": 1.4398,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001936931076244616,
+      "loss": 1.5191,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001936615077391985,
+      "loss": 1.5062,
+      "step": 3915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019362983147879826,
+      "loss": 1.5024,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019359807886909093,
+      "loss": 1.4988,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019356624993596878,
+      "loss": 1.4855,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019353434470538629,
+      "loss": 1.5001,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019350236320336023,
+      "loss": 1.4953,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001934703054559695,
+      "loss": 1.4843,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001934381714893552,
+      "loss": 1.4679,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019340596132972062,
+      "loss": 1.4638,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001933736750033311,
+      "loss": 1.465,
+      "step": 3960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019334131253651414,
+      "loss": 1.4728,
+      "step": 3965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019330887395565936,
+      "loss": 1.4941,
+      "step": 3970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019327635928721834,
+      "loss": 1.4869,
+      "step": 3975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019324376855770484,
+      "loss": 1.5167,
+      "step": 3980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019321110179369448,
+      "loss": 1.4614,
+      "step": 3985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019317835902182506,
+      "loss": 1.516,
+      "step": 3990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001931455402687963,
+      "loss": 1.4819,
+      "step": 3995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019311264556136975,
+      "loss": 1.4911,
+      "step": 4000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019307967492636905,
+      "loss": 1.4753,
+      "step": 4005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019304662839067974,
+      "loss": 1.487,
+      "step": 4010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019301350598124913,
+      "loss": 1.4955,
+      "step": 4015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019298030772508658,
+      "loss": 1.4849,
+      "step": 4020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019294703364926315,
+      "loss": 1.5116,
+      "step": 4025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019291368378091176,
+      "loss": 1.4563,
+      "step": 4030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001928802581472272,
+      "loss": 1.4931,
+      "step": 4035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000192846756775466,
+      "loss": 1.4745,
+      "step": 4040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019281317969294643,
+      "loss": 1.4956,
+      "step": 4045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019277952692704848,
+      "loss": 1.4962,
+      "step": 4050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019274579850521393,
+      "loss": 1.4451,
+      "step": 4055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019271199445494624,
+      "loss": 1.5035,
+      "step": 4060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019267811480381042,
+      "loss": 1.4918,
+      "step": 4065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001926441595794333,
+      "loss": 1.4766,
+      "step": 4070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019261012880950323,
+      "loss": 1.4601,
+      "step": 4075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019257602252177017,
+      "loss": 1.4964,
+      "step": 4080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019254184074404568,
+      "loss": 1.4682,
+      "step": 4085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001925075835042029,
+      "loss": 1.4707,
+      "step": 4090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019247325083017648,
+      "loss": 1.4259,
+      "step": 4095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019243884274996255,
+      "loss": 1.4643,
+      "step": 4100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019240435929161878,
+      "loss": 1.4715,
+      "step": 4105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019236980048326427,
+      "loss": 1.4774,
+      "step": 4110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001923351663530796,
+      "loss": 1.4701,
+      "step": 4115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019230045692930677,
+      "loss": 1.4838,
+      "step": 4120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019226567224024912,
+      "loss": 1.4794,
+      "step": 4125
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001922308123142714,
+      "loss": 1.4795,
+      "step": 4130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019219587717979973,
+      "loss": 1.4763,
+      "step": 4135
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019216086686532153,
+      "loss": 1.4965,
+      "step": 4140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019212578139938554,
+      "loss": 1.5114,
+      "step": 4145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019209062081060178,
+      "loss": 1.4927,
+      "step": 4150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019205538512764156,
+      "loss": 1.4886,
+      "step": 4155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001920200743792373,
+      "loss": 1.4667,
+      "step": 4160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019198468859418278,
+      "loss": 1.4878,
+      "step": 4165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019194922780133293,
+      "loss": 1.5009,
+      "step": 4170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019191369202960378,
+      "loss": 1.4706,
+      "step": 4175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019187808130797254,
+      "loss": 1.4836,
+      "step": 4180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019184239566547755,
+      "loss": 1.4744,
+      "step": 4185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019180663513121825,
+      "loss": 1.4759,
+      "step": 4190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001917707997343551,
+      "loss": 1.4465,
+      "step": 4195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019173488950410968,
+      "loss": 1.4752,
+      "step": 4200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019169890446976454,
+      "loss": 1.4593,
+      "step": 4205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019166284466066319,
+      "loss": 1.4992,
+      "step": 4210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019162671010621024,
+      "loss": 1.4524,
+      "step": 4215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001915905008358711,
+      "loss": 1.4715,
+      "step": 4220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001915542168791722,
+      "loss": 1.4902,
+      "step": 4225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001915178582657009,
+      "loss": 1.4699,
+      "step": 4230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019148142502510533,
+      "loss": 1.4632,
+      "step": 4235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019144491718709456,
+      "loss": 1.4785,
+      "step": 4240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019140833478143847,
+      "loss": 1.4622,
+      "step": 4245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001913716778379677,
+      "loss": 1.4746,
+      "step": 4250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019133494638657374,
+      "loss": 1.4865,
+      "step": 4255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001912981404572088,
+      "loss": 1.4808,
+      "step": 4260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019126126007988585,
+      "loss": 1.5115,
+      "step": 4265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001912243052846785,
+      "loss": 1.4847,
+      "step": 4270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001911872761017211,
+      "loss": 1.4664,
+      "step": 4275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019115017256120866,
+      "loss": 1.4713,
+      "step": 4280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001911129946933968,
+      "loss": 1.4886,
+      "step": 4285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019107574252860178,
+      "loss": 1.4937,
+      "step": 4290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019103841609720043,
+      "loss": 1.433,
+      "step": 4295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001910010154296301,
+      "loss": 1.4664,
+      "step": 4300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001909635405563887,
+      "loss": 1.4946,
+      "step": 4305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001909259915080347,
+      "loss": 1.4688,
+      "step": 4310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000190888368315187,
+      "loss": 1.4631,
+      "step": 4315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000190850671008525,
+      "loss": 1.458,
+      "step": 4320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019081289961878848,
+      "loss": 1.4863,
+      "step": 4325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019077505417677764,
+      "loss": 1.4789,
+      "step": 4330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019073713471335312,
+      "loss": 1.4919,
+      "step": 4335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019069914125943586,
+      "loss": 1.4988,
+      "step": 4340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001906610738460072,
+      "loss": 1.5027,
+      "step": 4345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019062293250410873,
+      "loss": 1.4981,
+      "step": 4350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019058471726484232,
+      "loss": 1.4197,
+      "step": 4355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019054642815937012,
+      "loss": 1.4616,
+      "step": 4360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019050806521891456,
+      "loss": 1.4806,
+      "step": 4365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001904696284747582,
+      "loss": 1.4921,
+      "step": 4370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019043111795824383,
+      "loss": 1.4368,
+      "step": 4375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019039253370077436,
+      "loss": 1.4956,
+      "step": 4380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001903538757338129,
+      "loss": 1.4884,
+      "step": 4385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019031514408888257,
+      "loss": 1.5047,
+      "step": 4390
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019027633879756663,
+      "loss": 1.441,
+      "step": 4395
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001902374598915084,
+      "loss": 1.4652,
+      "step": 4400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001901985074024112,
+      "loss": 1.4759,
+      "step": 4405
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019015948136203836,
+      "loss": 1.5075,
+      "step": 4410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019012038180221322,
+      "loss": 1.5099,
+      "step": 4415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019008120875481897,
+      "loss": 1.4607,
+      "step": 4420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019004196225179886,
+      "loss": 1.4744,
+      "step": 4425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019000264232515594,
+      "loss": 1.488,
+      "step": 4430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018996324900695318,
+      "loss": 1.462,
+      "step": 4435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001899237823293134,
+      "loss": 1.4705,
+      "step": 4440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018988424232441918,
+      "loss": 1.4757,
+      "step": 4445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000189844629024513,
+      "loss": 1.4953,
+      "step": 4450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018980494246189698,
+      "loss": 1.5236,
+      "step": 4455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018976518266893304,
+      "loss": 1.4874,
+      "step": 4460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018972534967804286,
+      "loss": 1.4537,
+      "step": 4465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018968544352170776,
+      "loss": 1.4815,
+      "step": 4470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018964546423246871,
+      "loss": 1.4856,
+      "step": 4475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018960541184292638,
+      "loss": 1.4881,
+      "step": 4480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018956528638574096,
+      "loss": 1.4695,
+      "step": 4485
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018952508789363227,
+      "loss": 1.4625,
+      "step": 4490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001894848163993797,
+      "loss": 1.475,
+      "step": 4495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018944447193582217,
+      "loss": 1.4838,
+      "step": 4500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018940405453585798,
+      "loss": 1.4659,
+      "step": 4505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018936356423244512,
+      "loss": 1.5054,
+      "step": 4510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001893230010586009,
+      "loss": 1.4518,
+      "step": 4515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5,
+      "learning_rate": 0.000189282365047402,
+      "loss": 1.4761,
+      "step": 4520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018924165623198462,
+      "loss": 1.4798,
+      "step": 4525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018920087464554427,
+      "loss": 1.4675,
+      "step": 4530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018916002032133574,
+      "loss": 1.4589,
+      "step": 4535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018911909329267325,
+      "loss": 1.4513,
+      "step": 4540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018907809359293025,
+      "loss": 1.4666,
+      "step": 4545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001890370212555394,
+      "loss": 1.4913,
+      "step": 4550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018899587631399266,
+      "loss": 1.4915,
+      "step": 4555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001889546588018412,
+      "loss": 1.476,
+      "step": 4560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001889133687526953,
+      "loss": 1.4865,
+      "step": 4565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018887200620022442,
+      "loss": 1.4912,
+      "step": 4570
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001888305711781572,
+      "loss": 1.493,
+      "step": 4575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001887890637202813,
+      "loss": 1.5079,
+      "step": 4580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018874748386044345,
+      "loss": 1.448,
+      "step": 4585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018870583163254948,
+      "loss": 1.496,
+      "step": 4590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018866410707056417,
+      "loss": 1.5013,
+      "step": 4595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001886223102085113,
+      "loss": 1.4647,
+      "step": 4600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018858044108047365,
+      "loss": 1.4696,
+      "step": 4605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018853849972059282,
+      "loss": 1.4539,
+      "step": 4610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018849648616306943,
+      "loss": 1.4826,
+      "step": 4615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018845440044216294,
+      "loss": 1.4955,
+      "step": 4620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001884122425921916,
+      "loss": 1.4647,
+      "step": 4625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018837001264753256,
+      "loss": 1.4642,
+      "step": 4630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018832771064262167,
+      "loss": 1.5195,
+      "step": 4635
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001882853366119536,
+      "loss": 1.4337,
+      "step": 4640
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018824289059008175,
+      "loss": 1.4322,
+      "step": 4645
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001882003726116182,
+      "loss": 1.4621,
+      "step": 4650
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018815778271123374,
+      "loss": 1.4737,
+      "step": 4655
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018811512092365776,
+      "loss": 1.481,
+      "step": 4660
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018807238728367828,
+      "loss": 1.4978,
+      "step": 4665
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000188029581826142,
+      "loss": 1.491,
+      "step": 4670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018798670458595402,
+      "loss": 1.4686,
+      "step": 4675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001879437555980781,
+      "loss": 1.4545,
+      "step": 4680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001879007348975365,
+      "loss": 1.4859,
+      "step": 4685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001878576425194099,
+      "loss": 1.4621,
+      "step": 4690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018781447849883744,
+      "loss": 1.502,
+      "step": 4695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018777124287101672,
+      "loss": 1.4946,
+      "step": 4700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001877279356712037,
+      "loss": 1.4527,
+      "step": 4705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018768455693471273,
+      "loss": 1.4475,
+      "step": 4710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001876411066969164,
+      "loss": 1.4761,
+      "step": 4715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018759758499324578,
+      "loss": 1.4905,
+      "step": 4720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018755399185919002,
+      "loss": 1.4484,
+      "step": 4725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001875103273302967,
+      "loss": 1.4452,
+      "step": 4730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018746659144217148,
+      "loss": 1.4934,
+      "step": 4735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018742278423047824,
+      "loss": 1.4497,
+      "step": 4740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018737890573093907,
+      "loss": 1.4335,
+      "step": 4745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018733495597933412,
+      "loss": 1.4861,
+      "step": 4750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018729093501150174,
+      "loss": 1.4859,
+      "step": 4755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018724684286333822,
+      "loss": 1.4934,
+      "step": 4760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018720267957079805,
+      "loss": 1.4894,
+      "step": 4765
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001871584451698936,
+      "loss": 1.4395,
+      "step": 4770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018711413969669526,
+      "loss": 1.4548,
+      "step": 4775
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018706976318733141,
+      "loss": 1.486,
+      "step": 4780
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018702531567798837,
+      "loss": 1.4525,
+      "step": 4785
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018698079720491024,
+      "loss": 1.5032,
+      "step": 4790
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018693620780439916,
+      "loss": 1.4398,
+      "step": 4795
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018689154751281494,
+      "loss": 1.4601,
+      "step": 4800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018684681636657529,
+      "loss": 1.4332,
+      "step": 4805
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001868020144021557,
+      "loss": 1.4677,
+      "step": 4810
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018675714165608935,
+      "loss": 1.445,
+      "step": 4815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018671219816496722,
+      "loss": 1.4806,
+      "step": 4820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018666718396543792,
+      "loss": 1.4517,
+      "step": 4825
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018662209909420772,
+      "loss": 1.4778,
+      "step": 4830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001865769435880405,
+      "loss": 1.4846,
+      "step": 4835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018653171748375785,
+      "loss": 1.4785,
+      "step": 4840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018648642081823877,
+      "loss": 1.4718,
+      "step": 4845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001864410536284199,
+      "loss": 1.4274,
+      "step": 4850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018639561595129537,
+      "loss": 1.5004,
+      "step": 4855
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001863501078239168,
+      "loss": 1.4704,
+      "step": 4860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001863045292833932,
+      "loss": 1.4659,
+      "step": 4865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018625888036689103,
+      "loss": 1.4741,
+      "step": 4870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001862131611116342,
+      "loss": 1.5111,
+      "step": 4875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001861673715549039,
+      "loss": 1.4657,
+      "step": 4880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001861215117340386,
+      "loss": 1.4569,
+      "step": 4885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018607558168643422,
+      "loss": 1.4685,
+      "step": 4890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001860295814495438,
+      "loss": 1.4891,
+      "step": 4895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018598351106087772,
+      "loss": 1.4572,
+      "step": 4900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001859373705580035,
+      "loss": 1.4894,
+      "step": 4905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018589115997854586,
+      "loss": 1.4507,
+      "step": 4910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018584487936018661,
+      "loss": 1.4602,
+      "step": 4915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018579852874066476,
+      "loss": 1.4457,
+      "step": 4920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001857521081577764,
+      "loss": 1.462,
+      "step": 4925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001857056176493745,
+      "loss": 1.4695,
+      "step": 4930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018565905725336933,
+      "loss": 1.4426,
+      "step": 4935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018561242700772788,
+      "loss": 1.5005,
+      "step": 4940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018556572695047427,
+      "loss": 1.4763,
+      "step": 4945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001855189571196895,
+      "loss": 1.4547,
+      "step": 4950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018547211755351147,
+      "loss": 1.4776,
+      "step": 4955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001854252082901349,
+      "loss": 1.4449,
+      "step": 4960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018537822936781132,
+      "loss": 1.4937,
+      "step": 4965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018533118082484927,
+      "loss": 1.4612,
+      "step": 4970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001852840626996138,
+      "loss": 1.4653,
+      "step": 4975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018523687503052685,
+      "loss": 1.4569,
+      "step": 4980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018518961785606703,
+      "loss": 1.4476,
+      "step": 4985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018514229121476962,
+      "loss": 1.4912,
+      "step": 4990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018509489514522657,
+      "loss": 1.4891,
+      "step": 4995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018504742968608639,
+      "loss": 1.4496,
+      "step": 5000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018499989487605423,
+      "loss": 1.4243,
+      "step": 5005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018495229075389183,
+      "loss": 1.4679,
+      "step": 5010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018490461735841732,
+      "loss": 1.4744,
+      "step": 5015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001848568747285054,
+      "loss": 1.483,
+      "step": 5020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018480906290308722,
+      "loss": 1.4682,
+      "step": 5025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018476118192115037,
+      "loss": 1.4677,
+      "step": 5030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018471323182173884,
+      "loss": 1.4572,
+      "step": 5035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018466521264395288,
+      "loss": 1.4501,
+      "step": 5040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001846171244269492,
+      "loss": 1.486,
+      "step": 5045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018456896720994072,
+      "loss": 1.484,
+      "step": 5050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001845207410321967,
+      "loss": 1.4618,
+      "step": 5055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018447244593304253,
+      "loss": 1.4758,
+      "step": 5060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001844240819518599,
+      "loss": 1.4699,
+      "step": 5065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018437564912808665,
+      "loss": 1.4397,
+      "step": 5070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001843271475012167,
+      "loss": 1.4299,
+      "step": 5075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018427857711080013,
+      "loss": 1.4455,
+      "step": 5080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018422993799644302,
+      "loss": 1.5103,
+      "step": 5085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018418123019780765,
+      "loss": 1.5017,
+      "step": 5090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001841324537546121,
+      "loss": 1.481,
+      "step": 5095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018408360870663063,
+      "loss": 1.4721,
+      "step": 5100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001840346950936932,
+      "loss": 1.4436,
+      "step": 5105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018398571295568595,
+      "loss": 1.4581,
+      "step": 5110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018393666233255073,
+      "loss": 1.4752,
+      "step": 5115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018388754326428524,
+      "loss": 1.5162,
+      "step": 5120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018383835579094304,
+      "loss": 1.4762,
+      "step": 5125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001837890999526335,
+      "loss": 1.431,
+      "step": 5130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001837397757895216,
+      "loss": 1.4589,
+      "step": 5135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018369038334182825,
+      "loss": 1.4569,
+      "step": 5140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018364092264982985,
+      "loss": 1.4379,
+      "step": 5145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018359139375385852,
+      "loss": 1.4665,
+      "step": 5150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000183541796694302,
+      "loss": 1.4636,
+      "step": 5155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018349213151160366,
+      "loss": 1.4701,
+      "step": 5160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018344239824626227,
+      "loss": 1.4937,
+      "step": 5165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001833925969388323,
+      "loss": 1.5028,
+      "step": 5170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018334272762992354,
+      "loss": 1.4863,
+      "step": 5175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001832927903602014,
+      "loss": 1.4618,
+      "step": 5180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001832427851703866,
+      "loss": 1.5137,
+      "step": 5185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018319271210125523,
+      "loss": 1.4636,
+      "step": 5190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018314257119363876,
+      "loss": 1.4666,
+      "step": 5195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018309236248842403,
+      "loss": 1.4487,
+      "step": 5200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018304208602655306,
+      "loss": 1.4146,
+      "step": 5205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018299174184902323,
+      "loss": 1.442,
+      "step": 5210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001829413299968871,
+      "loss": 1.4594,
+      "step": 5215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018289085051125233,
+      "loss": 1.4576,
+      "step": 5220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018284030343328181,
+      "loss": 1.4952,
+      "step": 5225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018278968880419363,
+      "loss": 1.4293,
+      "step": 5230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018273900666526078,
+      "loss": 1.4402,
+      "step": 5235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018268825705781145,
+      "loss": 1.4513,
+      "step": 5240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018263744002322874,
+      "loss": 1.4658,
+      "step": 5245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018258655560295087,
+      "loss": 1.4694,
+      "step": 5250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018253560383847082,
+      "loss": 1.4721,
+      "step": 5255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018248458477133662,
+      "loss": 1.4458,
+      "step": 5260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018243349844315117,
+      "loss": 1.4631,
+      "step": 5265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018238234489557215,
+      "loss": 1.4368,
+      "step": 5270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001823311241703122,
+      "loss": 1.4617,
+      "step": 5275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001822798363091385,
+      "loss": 1.4589,
+      "step": 5280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018222848135387323,
+      "loss": 1.4841,
+      "step": 5285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001821770593463931,
+      "loss": 1.4629,
+      "step": 5290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018212557032862953,
+      "loss": 1.4797,
+      "step": 5295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001820740143425687,
+      "loss": 1.4725,
+      "step": 5300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018202239143025125,
+      "loss": 1.4393,
+      "step": 5305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018197070163377248,
+      "loss": 1.4215,
+      "step": 5310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001819189449952822,
+      "loss": 1.5017,
+      "step": 5315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018186712155698475,
+      "loss": 1.4348,
+      "step": 5320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001818152313611389,
+      "loss": 1.4518,
+      "step": 5325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018176327445005788,
+      "loss": 1.4392,
+      "step": 5330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001817112508661093,
+      "loss": 1.4806,
+      "step": 5335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001816591606517152,
+      "loss": 1.4508,
+      "step": 5340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001816070038493519,
+      "loss": 1.4718,
+      "step": 5345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018155478050155,
+      "loss": 1.4973,
+      "step": 5350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018150249065089445,
+      "loss": 1.466,
+      "step": 5355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018145013434002434,
+      "loss": 1.4631,
+      "step": 5360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018139771161163295,
+      "loss": 1.4552,
+      "step": 5365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018134522250846783,
+      "loss": 1.4604,
+      "step": 5370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018129266707333052,
+      "loss": 1.4751,
+      "step": 5375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018124004534907675,
+      "loss": 1.4653,
+      "step": 5380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018118735737861625,
+      "loss": 1.4767,
+      "step": 5385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018113460320491278,
+      "loss": 1.4715,
+      "step": 5390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001810817828709841,
+      "loss": 1.4591,
+      "step": 5395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001810288964199019,
+      "loss": 1.458,
+      "step": 5400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018097594389479178,
+      "loss": 1.4588,
+      "step": 5405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00018092292533883325,
+      "loss": 1.4609,
+      "step": 5410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018086984079525965,
+      "loss": 1.4566,
+      "step": 5415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018081669030735814,
+      "loss": 1.4482,
+      "step": 5420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001807634739184696,
+      "loss": 1.4717,
+      "step": 5425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018071019167198872,
+      "loss": 1.4687,
+      "step": 5430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001806568436113638,
+      "loss": 1.4822,
+      "step": 5435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018060342978009697,
+      "loss": 1.489,
+      "step": 5440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018054995022174377,
+      "loss": 1.4717,
+      "step": 5445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018049640497991355,
+      "loss": 1.4647,
+      "step": 5450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001804427940982691,
+      "loss": 1.4401,
+      "step": 5455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018038911762052675,
+      "loss": 1.4547,
+      "step": 5460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018033537559045633,
+      "loss": 1.4169,
+      "step": 5465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018028156805188113,
+      "loss": 1.4402,
+      "step": 5470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018022769504867788,
+      "loss": 1.3978,
+      "step": 5475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018017375662477658,
+      "loss": 1.4334,
+      "step": 5480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018011975282416077,
+      "loss": 1.5001,
+      "step": 5485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018006568369086708,
+      "loss": 1.471,
+      "step": 5490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018001154926898565,
+      "loss": 1.472,
+      "step": 5495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017995734960265963,
+      "loss": 1.4583,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017990308473608555,
+      "loss": 1.4341,
+      "step": 5505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017984875471351302,
+      "loss": 1.4571,
+      "step": 5510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017979435957924476,
+      "loss": 1.4274,
+      "step": 5515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017973989937763665,
+      "loss": 1.436,
+      "step": 5520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001796853741530976,
+      "loss": 1.4752,
+      "step": 5525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001796307839500895,
+      "loss": 1.4609,
+      "step": 5530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017957612881312732,
+      "loss": 1.4496,
+      "step": 5535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017952140878677895,
+      "loss": 1.4494,
+      "step": 5540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001794666239156651,
+      "loss": 1.4503,
+      "step": 5545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017941177424445943,
+      "loss": 1.4531,
+      "step": 5550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017935685981788847,
+      "loss": 1.4794,
+      "step": 5555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017930188068073153,
+      "loss": 1.4549,
+      "step": 5560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017924683687782066,
+      "loss": 1.4862,
+      "step": 5565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017919172845404067,
+      "loss": 1.4818,
+      "step": 5570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017913655545432903,
+      "loss": 1.4655,
+      "step": 5575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017908131792367587,
+      "loss": 1.4568,
+      "step": 5580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017902601590712408,
+      "loss": 1.4522,
+      "step": 5585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017897064944976887,
+      "loss": 1.4274,
+      "step": 5590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017891521859675824,
+      "loss": 1.4413,
+      "step": 5595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017885972339329255,
+      "loss": 1.4872,
+      "step": 5600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017880416388462472,
+      "loss": 1.4727,
+      "step": 5605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001787485401160601,
+      "loss": 1.4454,
+      "step": 5610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017869285213295634,
+      "loss": 1.4622,
+      "step": 5615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001786370999807236,
+      "loss": 1.4323,
+      "step": 5620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017858128370482426,
+      "loss": 1.4826,
+      "step": 5625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017852540335077302,
+      "loss": 1.468,
+      "step": 5630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017846945896413685,
+      "loss": 1.4539,
+      "step": 5635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017841345059053492,
+      "loss": 1.4769,
+      "step": 5640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017835737827563857,
+      "loss": 1.4582,
+      "step": 5645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017830124206517128,
+      "loss": 1.4439,
+      "step": 5650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017824504200490866,
+      "loss": 1.4926,
+      "step": 5655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017818877814067833,
+      "loss": 1.4687,
+      "step": 5660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017813245051836,
+      "loss": 1.4605,
+      "step": 5665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017807605918388538,
+      "loss": 1.4395,
+      "step": 5670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017801960418323802,
+      "loss": 1.4392,
+      "step": 5675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001779630855624535,
+      "loss": 1.4397,
+      "step": 5680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017790650336761926,
+      "loss": 1.4667,
+      "step": 5685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017784985764487455,
+      "loss": 1.4608,
+      "step": 5690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017779314844041047,
+      "loss": 1.4725,
+      "step": 5695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017773637580046974,
+      "loss": 1.4486,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017767953977134704,
+      "loss": 1.4713,
+      "step": 5705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017762264039938855,
+      "loss": 1.4404,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001775656777309922,
+      "loss": 1.4726,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001775086518126075,
+      "loss": 1.472,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017745156269073555,
+      "loss": 1.412,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017739441041192896,
+      "loss": 1.4437,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017733719502279185,
+      "loss": 1.4726,
+      "step": 5735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017727991656997985,
+      "loss": 1.4665,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017722257510019996,
+      "loss": 1.4804,
+      "step": 5745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017716517066021056,
+      "loss": 1.4558,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017710770329682144,
+      "loss": 1.4385,
+      "step": 5755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017705017305689365,
+      "loss": 1.4651,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017699257998733952,
+      "loss": 1.4543,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017693492413512263,
+      "loss": 1.4433,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017687720554725772,
+      "loss": 1.4974,
+      "step": 5775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017681942427081072,
+      "loss": 1.4828,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017676158035289868,
+      "loss": 1.4557,
+      "step": 5785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00017670367384068971,
+      "loss": 1.4556,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017664570478140296,
+      "loss": 1.4639,
+      "step": 5795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017658767322230862,
+      "loss": 1.4575,
+      "step": 5800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017652957921072783,
+      "loss": 1.4221,
+      "step": 5805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001764714227940326,
+      "loss": 1.4913,
+      "step": 5810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001764132040196459,
+      "loss": 1.477,
+      "step": 5815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001763549229350415,
+      "loss": 1.4424,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017629657958774403,
+      "loss": 1.4894,
+      "step": 5825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017623817402532884,
+      "loss": 1.4499,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017617970629542207,
+      "loss": 1.4797,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017612117644570047,
+      "loss": 1.4671,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001760625845238915,
+      "loss": 1.4581,
+      "step": 5845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001760039305777733,
+      "loss": 1.4995,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001759452146551744,
+      "loss": 1.4475,
+      "step": 5855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017588643680397408,
+      "loss": 1.4706,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.921875,
+      "learning_rate": 0.00017582759707210203,
+      "loss": 1.4339,
+      "step": 5865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001757686955075383,
+      "loss": 1.4899,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017570973215831357,
+      "loss": 1.4016,
+      "step": 5875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017565070707250868,
+      "loss": 1.4766,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000175591620298255,
+      "loss": 1.4583,
+      "step": 5885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017553247188373402,
+      "loss": 1.4627,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017547326187717773,
+      "loss": 1.476,
+      "step": 5895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017541399032686811,
+      "loss": 1.4523,
+      "step": 5900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017535465728113746,
+      "loss": 1.4557,
+      "step": 5905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001752952627883682,
+      "loss": 1.4563,
+      "step": 5910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001752358068969928,
+      "loss": 1.4079,
+      "step": 5915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001751762896554939,
+      "loss": 1.4517,
+      "step": 5920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001751167111124041,
+      "loss": 1.4412,
+      "step": 5925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017505707131630597,
+      "loss": 1.4473,
+      "step": 5930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017499737031583207,
+      "loss": 1.4326,
+      "step": 5935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017493760815966486,
+      "loss": 1.4869,
+      "step": 5940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017487778489653667,
+      "loss": 1.4441,
+      "step": 5945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017481790057522964,
+      "loss": 1.4568,
+      "step": 5950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017475795524457568,
+      "loss": 1.4663,
+      "step": 5955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017469794895345656,
+      "loss": 1.4548,
+      "step": 5960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001746378817508036,
+      "loss": 1.4363,
+      "step": 5965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017457775368559793,
+      "loss": 1.463,
+      "step": 5970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017451756480687017,
+      "loss": 1.4627,
+      "step": 5975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001744573151637007,
+      "loss": 1.4662,
+      "step": 5980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00017439700480521934,
+      "loss": 1.4552,
+      "step": 5985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001743366337806054,
+      "loss": 1.4738,
+      "step": 5990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001742762021390877,
+      "loss": 1.4809,
+      "step": 5995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001742157099299445,
+      "loss": 1.4729,
+      "step": 6000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017415515720250346,
+      "loss": 1.4271,
+      "step": 6005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017409454400614153,
+      "loss": 1.4528,
+      "step": 6010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017403387039028503,
+      "loss": 1.4727,
+      "step": 6015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001739731364044095,
+      "loss": 1.4952,
+      "step": 6020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017391234209803975,
+      "loss": 1.4583,
+      "step": 6025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017385148752074975,
+      "loss": 1.4748,
+      "step": 6030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001737905727221626,
+      "loss": 1.4519,
+      "step": 6035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017372959775195057,
+      "loss": 1.4362,
+      "step": 6040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017366856265983493,
+      "loss": 1.4305,
+      "step": 6045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017360746749558602,
+      "loss": 1.4634,
+      "step": 6050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017354631230902316,
+      "loss": 1.4423,
+      "step": 6055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017348509715001457,
+      "loss": 1.3851,
+      "step": 6060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017342382206847744,
+      "loss": 1.4521,
+      "step": 6065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017336248711437774,
+      "loss": 1.4607,
+      "step": 6070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017330109233773037,
+      "loss": 1.4446,
+      "step": 6075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017323963778859892,
+      "loss": 1.4198,
+      "step": 6080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017317812351709576,
+      "loss": 1.4496,
+      "step": 6085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017311654957338196,
+      "loss": 1.4998,
+      "step": 6090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017305491600766725,
+      "loss": 1.4964,
+      "step": 6095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017299322287020995,
+      "loss": 1.467,
+      "step": 6100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017293147021131701,
+      "loss": 1.4281,
+      "step": 6105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017286965808134387,
+      "loss": 1.4406,
+      "step": 6110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017280778653069442,
+      "loss": 1.446,
+      "step": 6115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017274585560982117,
+      "loss": 1.4484,
+      "step": 6120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017268386536922487,
+      "loss": 1.457,
+      "step": 6125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017262181585945473,
+      "loss": 1.4568,
+      "step": 6130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017255970713110825,
+      "loss": 1.4554,
+      "step": 6135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017249753923483124,
+      "loss": 1.4656,
+      "step": 6140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017243531222131778,
+      "loss": 1.466,
+      "step": 6145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001723730261413101,
+      "loss": 1.4508,
+      "step": 6150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017231068104559864,
+      "loss": 1.4772,
+      "step": 6155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017224827698502195,
+      "loss": 1.4592,
+      "step": 6160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017218581401046666,
+      "loss": 1.4591,
+      "step": 6165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017212329217286743,
+      "loss": 1.4274,
+      "step": 6170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000172060711523207,
+      "loss": 1.4351,
+      "step": 6175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017199807211251588,
+      "loss": 1.4678,
+      "step": 6180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017193537399187272,
+      "loss": 1.4544,
+      "step": 6185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017187261721240388,
+      "loss": 1.4647,
+      "step": 6190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017180980182528364,
+      "loss": 1.4611,
+      "step": 6195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017174692788173403,
+      "loss": 1.4442,
+      "step": 6200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017168399543302486,
+      "loss": 1.4441,
+      "step": 6205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017162100453047363,
+      "loss": 1.4559,
+      "step": 6210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017155795522544548,
+      "loss": 1.449,
+      "step": 6215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001714948475693532,
+      "loss": 1.4631,
+      "step": 6220
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001714316816136572,
+      "loss": 1.4371,
+      "step": 6225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017136845740986533,
+      "loss": 1.4613,
+      "step": 6230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017130517500953306,
+      "loss": 1.4082,
+      "step": 6235
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001712418344642632,
+      "loss": 1.4541,
+      "step": 6240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017117843582570608,
+      "loss": 1.4396,
+      "step": 6245
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001711149791455593,
+      "loss": 1.4454,
+      "step": 6250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017105146447556787,
+      "loss": 1.4688,
+      "step": 6255
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017098789186752403,
+      "loss": 1.5107,
+      "step": 6260
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001709242613732673,
+      "loss": 1.4628,
+      "step": 6265
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001708605730446844,
+      "loss": 1.4343,
+      "step": 6270
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001707968269337092,
+      "loss": 1.4394,
+      "step": 6275
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017073302309232268,
+      "loss": 1.4945,
+      "step": 6280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017066916157255292,
+      "loss": 1.4309,
+      "step": 6285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017060524242647502,
+      "loss": 1.4539,
+      "step": 6290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017054126570621107,
+      "loss": 1.4287,
+      "step": 6295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017047723146393012,
+      "loss": 1.4452,
+      "step": 6300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017041313975184807,
+      "loss": 1.484,
+      "step": 6305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017034899062222776,
+      "loss": 1.4645,
+      "step": 6310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017028478412737882,
+      "loss": 1.4366,
+      "step": 6315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017022052031965762,
+      "loss": 1.4023,
+      "step": 6320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017015619925146735,
+      "loss": 1.4506,
+      "step": 6325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001700918209752578,
+      "loss": 1.4303,
+      "step": 6330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017002738554352552,
+      "loss": 1.4716,
+      "step": 6335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016996289300881353,
+      "loss": 1.4636,
+      "step": 6340
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016989834342371146,
+      "loss": 1.4566,
+      "step": 6345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016983373684085557,
+      "loss": 1.4486,
+      "step": 6350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016976907331292846,
+      "loss": 1.4181,
+      "step": 6355
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016970435289265923,
+      "loss": 1.4574,
+      "step": 6360
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016963957563282336,
+      "loss": 1.4554,
+      "step": 6365
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016957474158624266,
+      "loss": 1.4864,
+      "step": 6370
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001695098508057853,
+      "loss": 1.4319,
+      "step": 6375
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016944490334436566,
+      "loss": 1.4662,
+      "step": 6380
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016937989925494432,
+      "loss": 1.4459,
+      "step": 6385
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016931483859052813,
+      "loss": 1.4277,
+      "step": 6390
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016924972140417,
+      "loss": 1.4903,
+      "step": 6395
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00016918454774896892,
+      "loss": 1.4032,
+      "step": 6400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016911931767807,
+      "loss": 1.4108,
+      "step": 6405
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016905403124466427,
+      "loss": 1.466,
+      "step": 6410
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016898868850198878,
+      "loss": 1.4487,
+      "step": 6415
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001689232895033265,
+      "loss": 1.4416,
+      "step": 6420
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016885783430200616,
+      "loss": 1.4537,
+      "step": 6425
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001687923229514025,
+      "loss": 1.444,
+      "step": 6430
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016872675550493594,
+      "loss": 1.4532,
+      "step": 6435
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016866113201607257,
+      "loss": 1.4436,
+      "step": 6440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001685954525383244,
+      "loss": 1.4328,
+      "step": 6445
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001685297171252488,
+      "loss": 1.4651,
+      "step": 6450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001684639258304491,
+      "loss": 1.4652,
+      "step": 6455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016839807870757387,
+      "loss": 1.4639,
+      "step": 6460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016833217581031738,
+      "loss": 1.4381,
+      "step": 6465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016826621719241938,
+      "loss": 1.4469,
+      "step": 6470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00016820020290766498,
+      "loss": 1.434,
+      "step": 6475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016813413300988478,
+      "loss": 1.4403,
+      "step": 6480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001680680075529546,
+      "loss": 1.4353,
+      "step": 6485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016800182659079568,
+      "loss": 1.4958,
+      "step": 6490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001679355901773745,
+      "loss": 1.4549,
+      "step": 6495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001678692983667027,
+      "loss": 1.4271,
+      "step": 6500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016780295121283717,
+      "loss": 1.4726,
+      "step": 6505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016773654876987983,
+      "loss": 1.4884,
+      "step": 6510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016767009109197782,
+      "loss": 1.4783,
+      "step": 6515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016760357823332318,
+      "loss": 1.4433,
+      "step": 6520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00016753701024815304,
+      "loss": 1.4583,
+      "step": 6525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016747038719074945,
+      "loss": 1.4556,
+      "step": 6530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00016740370911543938,
+      "loss": 1.4543,
+      "step": 6535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016733697607659463,
+      "loss": 1.4557,
+      "step": 6540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001672701881286319,
+      "loss": 1.4402,
+      "step": 6545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016720334532601254,
+      "loss": 1.4782,
+      "step": 6550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016713644772324275,
+      "loss": 1.4437,
+      "step": 6555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016706949537487336,
+      "loss": 1.4701,
+      "step": 6560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001670024883354998,
+      "loss": 1.437,
+      "step": 6565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001669354266597622,
+      "loss": 1.4347,
+      "step": 6570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001668683104023452,
+      "loss": 1.463,
+      "step": 6575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016680113961797788,
+      "loss": 1.4704,
+      "step": 6580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016673391436143384,
+      "loss": 1.4307,
+      "step": 6585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016666663468753118,
+      "loss": 1.4539,
+      "step": 6590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016659930065113219,
+      "loss": 1.4447,
+      "step": 6595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016653191230714366,
+      "loss": 1.44,
+      "step": 6600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016646446971051653,
+      "loss": 1.4302,
+      "step": 6605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016639697291624615,
+      "loss": 1.4595,
+      "step": 6610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016632942197937185,
+      "loss": 1.4512,
+      "step": 6615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016626181695497726,
+      "loss": 1.4692,
+      "step": 6620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016619415789819012,
+      "loss": 1.4509,
+      "step": 6625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016612644486418211,
+      "loss": 1.47,
+      "step": 6630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016605867790816901,
+      "loss": 1.463,
+      "step": 6635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016599085708541065,
+      "loss": 1.423,
+      "step": 6640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001659229824512106,
+      "loss": 1.4258,
+      "step": 6645
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001658550540609164,
+      "loss": 1.4762,
+      "step": 6650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016578707196991953,
+      "loss": 1.4485,
+      "step": 6655
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016571903623365506,
+      "loss": 1.4438,
+      "step": 6660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016565094690760193,
+      "loss": 1.4411,
+      "step": 6665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016558280404728275,
+      "loss": 1.4442,
+      "step": 6670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016551460770826383,
+      "loss": 1.4678,
+      "step": 6675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016544635794615498,
+      "loss": 1.4506,
+      "step": 6680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00016537805481660968,
+      "loss": 1.4264,
+      "step": 6685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016530969837532487,
+      "loss": 1.4492,
+      "step": 6690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000165241288678041,
+      "loss": 1.4742,
+      "step": 6695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016517282578054187,
+      "loss": 1.454,
+      "step": 6700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001651043097386548,
+      "loss": 1.4484,
+      "step": 6705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001650357406082503,
+      "loss": 1.4599,
+      "step": 6710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016496711844524224,
+      "loss": 1.434,
+      "step": 6715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016489844330558773,
+      "loss": 1.4689,
+      "step": 6720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016482971524528714,
+      "loss": 1.4879,
+      "step": 6725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016476093432038385,
+      "loss": 1.4342,
+      "step": 6730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016469210058696446,
+      "loss": 1.4312,
+      "step": 6735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001646232141011586,
+      "loss": 1.4278,
+      "step": 6740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016455427491913888,
+      "loss": 1.4728,
+      "step": 6745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000164485283097121,
+      "loss": 1.4484,
+      "step": 6750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016441623869136343,
+      "loss": 1.4546,
+      "step": 6755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016434714175816764,
+      "loss": 1.4523,
+      "step": 6760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016427799235387784,
+      "loss": 1.4494,
+      "step": 6765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00016420879053488107,
+      "loss": 1.4351,
+      "step": 6770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016413953635760714,
+      "loss": 1.4547,
+      "step": 6775
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001640702298785285,
+      "loss": 1.4584,
+      "step": 6780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016400087115416034,
+      "loss": 1.4429,
+      "step": 6785
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001639314602410603,
+      "loss": 1.4234,
+      "step": 6790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016386199719582874,
+      "loss": 1.4738,
+      "step": 6795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001637924820751084,
+      "loss": 1.4311,
+      "step": 6800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016372291493558453,
+      "loss": 1.4664,
+      "step": 6805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016365329583398487,
+      "loss": 1.4486,
+      "step": 6810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016358362482707942,
+      "loss": 1.4347,
+      "step": 6815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001635139019716806,
+      "loss": 1.3988,
+      "step": 6820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000163444127324643,
+      "loss": 1.4205,
+      "step": 6825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016337430094286358,
+      "loss": 1.4269,
+      "step": 6830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016330442288328134,
+      "loss": 1.4432,
+      "step": 6835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016323449320287755,
+      "loss": 1.4516,
+      "step": 6840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001631645119586755,
+      "loss": 1.4376,
+      "step": 6845
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001630944792077405,
+      "loss": 1.457,
+      "step": 6850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016302439500718002,
+      "loss": 1.472,
+      "step": 6855
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016295425941414323,
+      "loss": 1.4462,
+      "step": 6860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00016288407248582146,
+      "loss": 1.4394,
+      "step": 6865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001628138342794477,
+      "loss": 1.4604,
+      "step": 6870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016274354485229688,
+      "loss": 1.4599,
+      "step": 6875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001626732042616857,
+      "loss": 1.4046,
+      "step": 6880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016260281256497247,
+      "loss": 1.4412,
+      "step": 6885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016253236981955726,
+      "loss": 1.4617,
+      "step": 6890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016246187608288178,
+      "loss": 1.4696,
+      "step": 6895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016239133141242925,
+      "loss": 1.4534,
+      "step": 6900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001623207358657245,
+      "loss": 1.4349,
+      "step": 6905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001622500895003338,
+      "loss": 1.4505,
+      "step": 6910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016217939237386485,
+      "loss": 1.432,
+      "step": 6915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016210864454396678,
+      "loss": 1.4469,
+      "step": 6920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016203784606833,
+      "loss": 1.4463,
+      "step": 6925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016196699700468634,
+      "loss": 1.4452,
+      "step": 6930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001618960974108088,
+      "loss": 1.4029,
+      "step": 6935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001618251473445115,
+      "loss": 1.4377,
+      "step": 6940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016175414686364994,
+      "loss": 1.4317,
+      "step": 6945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016168309602612052,
+      "loss": 1.4971,
+      "step": 6950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016161199488986077,
+      "loss": 1.4458,
+      "step": 6955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016154084351284925,
+      "loss": 1.4253,
+      "step": 6960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00016146964195310555,
+      "loss": 1.4211,
+      "step": 6965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016139839026869005,
+      "loss": 1.4844,
+      "step": 6970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016132708851770408,
+      "loss": 1.4599,
+      "step": 6975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016125573675828983,
+      "loss": 1.4546,
+      "step": 6980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016118433504863012,
+      "loss": 1.4241,
+      "step": 6985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016111288344694875,
+      "loss": 1.4573,
+      "step": 6990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016104138201150994,
+      "loss": 1.4118,
+      "step": 6995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016096983080061874,
+      "loss": 1.4024,
+      "step": 7000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00016089822987262067,
+      "loss": 1.4301,
+      "step": 7005
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016082657928590183,
+      "loss": 1.4426,
+      "step": 7010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016075487909888886,
+      "loss": 1.4379,
+      "step": 7015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001606831293700488,
+      "loss": 1.4465,
+      "step": 7020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016061133015788905,
+      "loss": 1.466,
+      "step": 7025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016053948152095745,
+      "loss": 1.4491,
+      "step": 7030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001604675835178421,
+      "loss": 1.4804,
+      "step": 7035
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016039563620717128,
+      "loss": 1.4269,
+      "step": 7040
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016032363964761363,
+      "loss": 1.4239,
+      "step": 7045
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016025159389787788,
+      "loss": 1.4139,
+      "step": 7050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016017949901671276,
+      "loss": 1.444,
+      "step": 7055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016010735506290726,
+      "loss": 1.4524,
+      "step": 7060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016003516209529023,
+      "loss": 1.4677,
+      "step": 7065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015996292017273058,
+      "loss": 1.4294,
+      "step": 7070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001598906293541371,
+      "loss": 1.4313,
+      "step": 7075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015981828969845844,
+      "loss": 1.4712,
+      "step": 7080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015974590126468315,
+      "loss": 1.4572,
+      "step": 7085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00015967346411183941,
+      "loss": 1.4446,
+      "step": 7090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015960097829899528,
+      "loss": 1.4755,
+      "step": 7095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001595284438852584,
+      "loss": 1.4535,
+      "step": 7100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015945586092977612,
+      "loss": 1.4407,
+      "step": 7105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015938322949173527,
+      "loss": 1.4454,
+      "step": 7110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015931054963036232,
+      "loss": 1.4415,
+      "step": 7115
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015923782140492317,
+      "loss": 1.4375,
+      "step": 7120
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015916504487472314,
+      "loss": 1.4466,
+      "step": 7125
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000159092220099107,
+      "loss": 1.4097,
+      "step": 7130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001590193471374588,
+      "loss": 1.4065,
+      "step": 7135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015894642604920192,
+      "loss": 1.4808,
+      "step": 7140
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015887345689379897,
+      "loss": 1.4544,
+      "step": 7145
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015880043973075177,
+      "loss": 1.4441,
+      "step": 7150
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015872737461960126,
+      "loss": 1.4611,
+      "step": 7155
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015865426161992753,
+      "loss": 1.4104,
+      "step": 7160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015858110079134966,
+      "loss": 1.4313,
+      "step": 7165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015850789219352577,
+      "loss": 1.4332,
+      "step": 7170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001584346358861529,
+      "loss": 1.4422,
+      "step": 7175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015836133192896702,
+      "loss": 1.4478,
+      "step": 7180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00015828798038174298,
+      "loss": 1.4422,
+      "step": 7185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001582145813042944,
+      "loss": 1.4665,
+      "step": 7190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001581411347564736,
+      "loss": 1.4325,
+      "step": 7195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015806764079817178,
+      "loss": 1.3919,
+      "step": 7200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001579940994893186,
+      "loss": 1.4532,
+      "step": 7205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015792051088988246,
+      "loss": 1.4209,
+      "step": 7210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00015784687505987033,
+      "loss": 1.4561,
+      "step": 7215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015777319205932758,
+      "loss": 1.442,
+      "step": 7220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00015769946194833817,
+      "loss": 1.4268,
+      "step": 7225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015762568478702436,
+      "loss": 1.4749,
+      "step": 7230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015755186063554696,
+      "loss": 1.4237,
+      "step": 7235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015747798955410483,
+      "loss": 1.4823,
+      "step": 7240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015740407160293535,
+      "loss": 1.4482,
+      "step": 7245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00015733010684231395,
+      "loss": 1.4474,
+      "step": 7250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015725609533255434,
+      "loss": 1.4608,
+      "step": 7255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015718203713400828,
+      "loss": 1.4731,
+      "step": 7260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001571079323070656,
+      "loss": 1.412,
+      "step": 7265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015703378091215428,
+      "loss": 1.4288,
+      "step": 7270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015695958300974007,
+      "loss": 1.4882,
+      "step": 7275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001568853386603268,
+      "loss": 1.4557,
+      "step": 7280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001568110479244561,
+      "loss": 1.4651,
+      "step": 7285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015673671086270741,
+      "loss": 1.4711,
+      "step": 7290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015666232753569807,
+      "loss": 1.4877,
+      "step": 7295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000156587898004083,
+      "loss": 1.4273,
+      "step": 7300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015651342232855486,
+      "loss": 1.4321,
+      "step": 7305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015643890056984394,
+      "loss": 1.4391,
+      "step": 7310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015636433278871814,
+      "loss": 1.453,
+      "step": 7315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015628971904598277,
+      "loss": 1.4338,
+      "step": 7320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015621505940248076,
+      "loss": 1.4304,
+      "step": 7325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015614035391909242,
+      "loss": 1.4304,
+      "step": 7330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015606560265673535,
+      "loss": 1.4565,
+      "step": 7335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015599080567636463,
+      "loss": 1.465,
+      "step": 7340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015591596303897256,
+      "loss": 1.449,
+      "step": 7345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015584107480558858,
+      "loss": 1.4558,
+      "step": 7350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015576614103727946,
+      "loss": 1.464,
+      "step": 7355
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015569116179514896,
+      "loss": 1.4592,
+      "step": 7360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015561613714033804,
+      "loss": 1.4774,
+      "step": 7365
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00015554106713402466,
+      "loss": 1.4422,
+      "step": 7370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015546595183742372,
+      "loss": 1.3744,
+      "step": 7375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015539079131178705,
+      "loss": 1.4724,
+      "step": 7380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015531558561840343,
+      "loss": 1.4461,
+      "step": 7385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015524033481859842,
+      "loss": 1.4322,
+      "step": 7390
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015516503897373434,
+      "loss": 1.4249,
+      "step": 7395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015508969814521025,
+      "loss": 1.4321,
+      "step": 7400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015501431239446197,
+      "loss": 1.4365,
+      "step": 7405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015493888178296191,
+      "loss": 1.4282,
+      "step": 7410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015486340637221895,
+      "loss": 1.4296,
+      "step": 7415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015478788622377872,
+      "loss": 1.4462,
+      "step": 7420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015471232139922312,
+      "loss": 1.443,
+      "step": 7425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015463671196017055,
+      "loss": 1.4506,
+      "step": 7430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015456105796827588,
+      "loss": 1.4532,
+      "step": 7435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015448535948523018,
+      "loss": 1.45,
+      "step": 7440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015440961657276088,
+      "loss": 1.4341,
+      "step": 7445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001543338292926316,
+      "loss": 1.4292,
+      "step": 7450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001542579977066422,
+      "loss": 1.4081,
+      "step": 7455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015418212187662858,
+      "loss": 1.4163,
+      "step": 7460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015410620186446277,
+      "loss": 1.4281,
+      "step": 7465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015403023773205286,
+      "loss": 1.4514,
+      "step": 7470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015395422954134278,
+      "loss": 1.4534,
+      "step": 7475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001538781773543126,
+      "loss": 1.4488,
+      "step": 7480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001538020812329781,
+      "loss": 1.4453,
+      "step": 7485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015372594123939094,
+      "loss": 1.4087,
+      "step": 7490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015364975743563858,
+      "loss": 1.4497,
+      "step": 7495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001535735298838441,
+      "loss": 1.4217,
+      "step": 7500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015349725864616639,
+      "loss": 1.4277,
+      "step": 7505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015342094378479988,
+      "loss": 1.4472,
+      "step": 7510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001533445853619746,
+      "loss": 1.4571,
+      "step": 7515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001532681834399561,
+      "loss": 1.4408,
+      "step": 7520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001531917380810454,
+      "loss": 1.4034,
+      "step": 7525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015311524934757893,
+      "loss": 1.4527,
+      "step": 7530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001530387173019285,
+      "loss": 1.4673,
+      "step": 7535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015296214200650126,
+      "loss": 1.4032,
+      "step": 7540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015288552352373956,
+      "loss": 1.4031,
+      "step": 7545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000152808861916121,
+      "loss": 1.4263,
+      "step": 7550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015273215724615846,
+      "loss": 1.4538,
+      "step": 7555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015265540957639973,
+      "loss": 1.4294,
+      "step": 7560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015257861896942777,
+      "loss": 1.4324,
+      "step": 7565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001525017854878606,
+      "loss": 1.4431,
+      "step": 7570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001524249091943511,
+      "loss": 1.4652,
+      "step": 7575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015234799015158713,
+      "loss": 1.446,
+      "step": 7580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015227102842229134,
+      "loss": 1.4417,
+      "step": 7585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001521940240692213,
+      "loss": 1.426,
+      "step": 7590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015211697715516927,
+      "loss": 1.4178,
+      "step": 7595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001520398877429622,
+      "loss": 1.4294,
+      "step": 7600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015196275589546168,
+      "loss": 1.442,
+      "step": 7605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000151885581675564,
+      "loss": 1.4626,
+      "step": 7610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001518083651461999,
+      "loss": 1.4439,
+      "step": 7615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001517311063703347,
+      "loss": 1.3871,
+      "step": 7620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015165380541096803,
+      "loss": 1.4404,
+      "step": 7625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015157646233113412,
+      "loss": 1.4486,
+      "step": 7630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001514990771939014,
+      "loss": 1.4766,
+      "step": 7635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015142165006237266,
+      "loss": 1.4596,
+      "step": 7640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001513441809996849,
+      "loss": 1.4263,
+      "step": 7645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001512666700690093,
+      "loss": 1.4117,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015118911733355125,
+      "loss": 1.4577,
+      "step": 7655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015111152285655013,
+      "loss": 1.4462,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015103388670127947,
+      "loss": 1.433,
+      "step": 7665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001509562089310467,
+      "loss": 1.4355,
+      "step": 7670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001508784896091932,
+      "loss": 1.3756,
+      "step": 7675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001508007287990943,
+      "loss": 1.4544,
+      "step": 7680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015072292656415906,
+      "loss": 1.4182,
+      "step": 7685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015064508296783037,
+      "loss": 1.4554,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00015056719807358485,
+      "loss": 1.4284,
+      "step": 7695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015048927194493276,
+      "loss": 1.4262,
+      "step": 7700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015041130464541808,
+      "loss": 1.4831,
+      "step": 7705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015033329623861822,
+      "loss": 1.4461,
+      "step": 7710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00015025524678814427,
+      "loss": 1.4351,
+      "step": 7715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015017715635764063,
+      "loss": 1.4615,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00015009902501078525,
+      "loss": 1.4451,
+      "step": 7725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001500208528112893,
+      "loss": 1.424,
+      "step": 7730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014994263982289746,
+      "loss": 1.4405,
+      "step": 7735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014986438610938748,
+      "loss": 1.4386,
+      "step": 7740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014978609173457044,
+      "loss": 1.4277,
+      "step": 7745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001497077567622905,
+      "loss": 1.4664,
+      "step": 7750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014962938125642503,
+      "loss": 1.4372,
+      "step": 7755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014955096528088428,
+      "loss": 1.4382,
+      "step": 7760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014947250889961168,
+      "loss": 1.4066,
+      "step": 7765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001493940121765835,
+      "loss": 1.4029,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014931547517580898,
+      "loss": 1.4455,
+      "step": 7775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014923689796133007,
+      "loss": 1.4802,
+      "step": 7780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001491582805972217,
+      "loss": 1.4362,
+      "step": 7785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014907962314759143,
+      "loss": 1.4446,
+      "step": 7790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014900092567657946,
+      "loss": 1.4295,
+      "step": 7795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014892218824835872,
+      "loss": 1.4239,
+      "step": 7800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001488434109271347,
+      "loss": 1.4603,
+      "step": 7805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014876459377714541,
+      "loss": 1.458,
+      "step": 7810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001486857368626613,
+      "loss": 1.4326,
+      "step": 7815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014860684024798536,
+      "loss": 1.4336,
+      "step": 7820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014852790399745276,
+      "loss": 1.4285,
+      "step": 7825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014844892817543118,
+      "loss": 1.444,
+      "step": 7830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014836991284632048,
+      "loss": 1.4552,
+      "step": 7835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014829085807455274,
+      "loss": 1.4585,
+      "step": 7840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014821176392459224,
+      "loss": 1.4027,
+      "step": 7845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001481326304609353,
+      "loss": 1.4385,
+      "step": 7850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001480534577481104,
+      "loss": 1.4407,
+      "step": 7855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014797424585067789,
+      "loss": 1.4317,
+      "step": 7860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001478949948332302,
+      "loss": 1.4478,
+      "step": 7865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014781570476039163,
+      "loss": 1.4349,
+      "step": 7870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014773637569681823,
+      "loss": 1.4618,
+      "step": 7875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014765700770719796,
+      "loss": 1.4413,
+      "step": 7880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014757760085625047,
+      "loss": 1.459,
+      "step": 7885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014749815520872717,
+      "loss": 1.4563,
+      "step": 7890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014741867082941095,
+      "loss": 1.461,
+      "step": 7895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014733914778311647,
+      "loss": 1.4266,
+      "step": 7900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014725958613468976,
+      "loss": 1.4292,
+      "step": 7905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014717998594900844,
+      "loss": 1.4546,
+      "step": 7910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001471003472909815,
+      "loss": 1.4307,
+      "step": 7915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001470206702255493,
+      "loss": 1.4127,
+      "step": 7920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014694095481768358,
+      "loss": 1.418,
+      "step": 7925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014686120113238725,
+      "loss": 1.4108,
+      "step": 7930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014678140923469452,
+      "loss": 1.4327,
+      "step": 7935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001467015791896707,
+      "loss": 1.4625,
+      "step": 7940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014662171106241223,
+      "loss": 1.4198,
+      "step": 7945
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001465418049180466,
+      "loss": 1.4784,
+      "step": 7950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014646186082173233,
+      "loss": 1.4499,
+      "step": 7955
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001463818788386588,
+      "loss": 1.4569,
+      "step": 7960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014630185903404642,
+      "loss": 1.4456,
+      "step": 7965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014622180147314632,
+      "loss": 1.4491,
+      "step": 7970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001461417062212405,
+      "loss": 1.464,
+      "step": 7975
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014606157334364162,
+      "loss": 1.4202,
+      "step": 7980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014598140290569307,
+      "loss": 1.4205,
+      "step": 7985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014590119497276887,
+      "loss": 1.4561,
+      "step": 7990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001458209496102736,
+      "loss": 1.4398,
+      "step": 7995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014574066688364235,
+      "loss": 1.4299,
+      "step": 8000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001456603468583407,
+      "loss": 1.4382,
+      "step": 8005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014557998959986466,
+      "loss": 1.4097,
+      "step": 8010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014549959517374056,
+      "loss": 1.4518,
+      "step": 8015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014541916364552504,
+      "loss": 1.438,
+      "step": 8020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014533869508080504,
+      "loss": 1.458,
+      "step": 8025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014525818954519765,
+      "loss": 1.4477,
+      "step": 8030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001451776471043502,
+      "loss": 1.4434,
+      "step": 8035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014509706782393992,
+      "loss": 1.4283,
+      "step": 8040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014501645176967428,
+      "loss": 1.452,
+      "step": 8045
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014493579900729065,
+      "loss": 1.4493,
+      "step": 8050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014485510960255638,
+      "loss": 1.4199,
+      "step": 8055
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001447743836212686,
+      "loss": 1.4594,
+      "step": 8060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014469362112925436,
+      "loss": 1.4306,
+      "step": 8065
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001446128221923704,
+      "loss": 1.4349,
+      "step": 8070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014453198687650336,
+      "loss": 1.459,
+      "step": 8075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001444511152475693,
+      "loss": 1.4249,
+      "step": 8080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014437020737151403,
+      "loss": 1.4304,
+      "step": 8085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014428926331431293,
+      "loss": 1.428,
+      "step": 8090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014420828314197078,
+      "loss": 1.4365,
+      "step": 8095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014412726692052195,
+      "loss": 1.4066,
+      "step": 8100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014404621471603005,
+      "loss": 1.4333,
+      "step": 8105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014396512659458824,
+      "loss": 1.4428,
+      "step": 8110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001438840026223187,
+      "loss": 1.4507,
+      "step": 8115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014380284286537307,
+      "loss": 1.4596,
+      "step": 8120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014372164738993206,
+      "loss": 1.4695,
+      "step": 8125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00014364041626220556,
+      "loss": 1.4101,
+      "step": 8130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014355914954843247,
+      "loss": 1.424,
+      "step": 8135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014347784731488078,
+      "loss": 1.448,
+      "step": 8140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014339650962784736,
+      "loss": 1.4513,
+      "step": 8145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014331513655365806,
+      "loss": 1.4363,
+      "step": 8150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014323372815866757,
+      "loss": 1.44,
+      "step": 8155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014315228450925943,
+      "loss": 1.469,
+      "step": 8160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014307080567184575,
+      "loss": 1.4107,
+      "step": 8165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014298929171286753,
+      "loss": 1.4158,
+      "step": 8170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014290774269879434,
+      "loss": 1.417,
+      "step": 8175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014282615869612433,
+      "loss": 1.45,
+      "step": 8180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014274453977138415,
+      "loss": 1.4191,
+      "step": 8185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000142662885991129,
+      "loss": 1.446,
+      "step": 8190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014258119742194242,
+      "loss": 1.4128,
+      "step": 8195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014249947413043642,
+      "loss": 1.4124,
+      "step": 8200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014241771618325123,
+      "loss": 1.4235,
+      "step": 8205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014233592364705535,
+      "loss": 1.4351,
+      "step": 8210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001422540965885455,
+      "loss": 1.4278,
+      "step": 8215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014217223507444662,
+      "loss": 1.4011,
+      "step": 8220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014209033917151167,
+      "loss": 1.4438,
+      "step": 8225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014200840894652167,
+      "loss": 1.4237,
+      "step": 8230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014192644446628556,
+      "loss": 1.3854,
+      "step": 8235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014184444579764036,
+      "loss": 1.4199,
+      "step": 8240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001417624130074508,
+      "loss": 1.4375,
+      "step": 8245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014168034616260963,
+      "loss": 1.4375,
+      "step": 8250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014159824533003718,
+      "loss": 1.4656,
+      "step": 8255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001415161105766816,
+      "loss": 1.4362,
+      "step": 8260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001414339419695187,
+      "loss": 1.4265,
+      "step": 8265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014135173957555182,
+      "loss": 1.4623,
+      "step": 8270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014126950346181195,
+      "loss": 1.4236,
+      "step": 8275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014118723369535747,
+      "loss": 1.4621,
+      "step": 8280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001411049303432743,
+      "loss": 1.4385,
+      "step": 8285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014102259347267574,
+      "loss": 1.4639,
+      "step": 8290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014094022315070236,
+      "loss": 1.4278,
+      "step": 8295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014085781944452201,
+      "loss": 1.426,
+      "step": 8300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001407753824213298,
+      "loss": 1.4421,
+      "step": 8305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014069291214834802,
+      "loss": 1.4385,
+      "step": 8310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014061040869282608,
+      "loss": 1.4284,
+      "step": 8315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014052787212204032,
+      "loss": 1.4321,
+      "step": 8320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014044530250329425,
+      "loss": 1.416,
+      "step": 8325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001403626999039183,
+      "loss": 1.4127,
+      "step": 8330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014028006439126967,
+      "loss": 1.4603,
+      "step": 8335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014019739603273251,
+      "loss": 1.4149,
+      "step": 8340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014011469489571776,
+      "loss": 1.4229,
+      "step": 8345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014003196104766304,
+      "loss": 1.4199,
+      "step": 8350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013994919455603263,
+      "loss": 1.4012,
+      "step": 8355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013986639548831752,
+      "loss": 1.4263,
+      "step": 8360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013978356391203514,
+      "loss": 1.4384,
+      "step": 8365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001397006998947295,
+      "loss": 1.4276,
+      "step": 8370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013961780350397112,
+      "loss": 1.4205,
+      "step": 8375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013953487480735679,
+      "loss": 1.4579,
+      "step": 8380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013945191387250972,
+      "loss": 1.4331,
+      "step": 8385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013936892076707937,
+      "loss": 1.4885,
+      "step": 8390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001392858955587415,
+      "loss": 1.397,
+      "step": 8395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013920283831519802,
+      "loss": 1.4128,
+      "step": 8400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001391197491041769,
+      "loss": 1.4026,
+      "step": 8405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013903662799343226,
+      "loss": 1.4581,
+      "step": 8410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013895347505074417,
+      "loss": 1.4506,
+      "step": 8415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001388702903439187,
+      "loss": 1.4552,
+      "step": 8420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013878707394078782,
+      "loss": 1.433,
+      "step": 8425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013870382590920933,
+      "loss": 1.4658,
+      "step": 8430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001386205463170668,
+      "loss": 1.4267,
+      "step": 8435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013853723523226955,
+      "loss": 1.4343,
+      "step": 8440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013845389272275268,
+      "loss": 1.4299,
+      "step": 8445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001383705188564767,
+      "loss": 1.4275,
+      "step": 8450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013828711370142792,
+      "loss": 1.4335,
+      "step": 8455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013820367732561803,
+      "loss": 1.4597,
+      "step": 8460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013812020979708418,
+      "loss": 1.422,
+      "step": 8465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013803671118388895,
+      "loss": 1.4521,
+      "step": 8470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001379531815541203,
+      "loss": 1.4216,
+      "step": 8475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013786962097589144,
+      "loss": 1.4213,
+      "step": 8480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001377860295173408,
+      "loss": 1.4642,
+      "step": 8485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00013770240724663208,
+      "loss": 1.397,
+      "step": 8490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013761875423195396,
+      "loss": 1.4256,
+      "step": 8495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013753507054152034,
+      "loss": 1.4535,
+      "step": 8500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013745135624357007,
+      "loss": 1.445,
+      "step": 8505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001373676114063669,
+      "loss": 1.4233,
+      "step": 8510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013728383609819958,
+      "loss": 1.4573,
+      "step": 8515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013720003038738163,
+      "loss": 1.4007,
+      "step": 8520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013711619434225145,
+      "loss": 1.4374,
+      "step": 8525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001370323280311721,
+      "loss": 1.4193,
+      "step": 8530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013694843152253132,
+      "loss": 1.4188,
+      "step": 8535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013686450488474154,
+      "loss": 1.4665,
+      "step": 8540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013678054818623965,
+      "loss": 1.4242,
+      "step": 8545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013669656149548718,
+      "loss": 1.4485,
+      "step": 8550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013661254488097003,
+      "loss": 1.4056,
+      "step": 8555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013652849841119856,
+      "loss": 1.4034,
+      "step": 8560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013644442215470737,
+      "loss": 1.4411,
+      "step": 8565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013636031618005553,
+      "loss": 1.4342,
+      "step": 8570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001362761805558261,
+      "loss": 1.4277,
+      "step": 8575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013619201535062657,
+      "loss": 1.4415,
+      "step": 8580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013610782063308837,
+      "loss": 1.4548,
+      "step": 8585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013602359647186708,
+      "loss": 1.4578,
+      "step": 8590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013593934293564222,
+      "loss": 1.4366,
+      "step": 8595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013585506009311738,
+      "loss": 1.4126,
+      "step": 8600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013577074801301992,
+      "loss": 1.4599,
+      "step": 8605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013568640676410115,
+      "loss": 1.3858,
+      "step": 8610
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013560203641513606,
+      "loss": 1.4707,
+      "step": 8615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001355176370349235,
+      "loss": 1.4366,
+      "step": 8620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013543320869228585,
+      "loss": 1.4617,
+      "step": 8625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013534875145606925,
+      "loss": 1.3993,
+      "step": 8630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013526426539514324,
+      "loss": 1.4384,
+      "step": 8635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013517975057840097,
+      "loss": 1.4434,
+      "step": 8640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00013509520707475907,
+      "loss": 1.4362,
+      "step": 8645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013501063495315743,
+      "loss": 1.4159,
+      "step": 8650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001349260342825595,
+      "loss": 1.3869,
+      "step": 8655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013484140513195166,
+      "loss": 1.3795,
+      "step": 8660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001347567475703439,
+      "loss": 1.3984,
+      "step": 8665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013467206166676914,
+      "loss": 1.4237,
+      "step": 8670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001345873474902835,
+      "loss": 1.4428,
+      "step": 8675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000134502605109966,
+      "loss": 1.4228,
+      "step": 8680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013441783459491893,
+      "loss": 1.4034,
+      "step": 8685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013433303601426727,
+      "loss": 1.4506,
+      "step": 8690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001342482094371591,
+      "loss": 1.4052,
+      "step": 8695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013416335493276511,
+      "loss": 1.4684,
+      "step": 8700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013407847257027896,
+      "loss": 1.3952,
+      "step": 8705
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013399356241891686,
+      "loss": 1.4229,
+      "step": 8710
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013390862454791785,
+      "loss": 1.4695,
+      "step": 8715
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013382365902654336,
+      "loss": 1.4372,
+      "step": 8720
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00013373866592407765,
+      "loss": 1.4514,
+      "step": 8725
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013365364530982716,
+      "loss": 1.4135,
+      "step": 8730
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013356859725312104,
+      "loss": 1.4251,
+      "step": 8735
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001334835218233106,
+      "loss": 1.4028,
+      "step": 8740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013339841908976963,
+      "loss": 1.3512,
+      "step": 8745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013331328912189407,
+      "loss": 1.4351,
+      "step": 8750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013322813198910212,
+      "loss": 1.4621,
+      "step": 8755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001331429477608342,
+      "loss": 1.4065,
+      "step": 8760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013305773650655267,
+      "loss": 1.4684,
+      "step": 8765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013297249829574202,
+      "loss": 1.4672,
+      "step": 8770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013288723319790875,
+      "loss": 1.4093,
+      "step": 8775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00013280194128258122,
+      "loss": 1.4403,
+      "step": 8780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013271662261930971,
+      "loss": 1.4036,
+      "step": 8785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013263127727766624,
+      "loss": 1.4239,
+      "step": 8790
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013254590532724468,
+      "loss": 1.4219,
+      "step": 8795
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013246050683766048,
+      "loss": 1.4247,
+      "step": 8800
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013237508187855093,
+      "loss": 1.4521,
+      "step": 8805
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001322896305195746,
+      "loss": 1.4264,
+      "step": 8810
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013220415283041195,
+      "loss": 1.4408,
+      "step": 8815
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013211864888076457,
+      "loss": 1.4116,
+      "step": 8820
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013203311874035567,
+      "loss": 1.4312,
+      "step": 8825
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013194756247892977,
+      "loss": 1.4617,
+      "step": 8830
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013186198016625268,
+      "loss": 1.4574,
+      "step": 8835
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013177637187211143,
+      "loss": 1.433,
+      "step": 8840
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013169073766631427,
+      "loss": 1.4582,
+      "step": 8845
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013160507761869063,
+      "loss": 1.4727,
+      "step": 8850
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013151939179909086,
+      "loss": 1.4673,
+      "step": 8855
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001314336802773865,
+      "loss": 1.4465,
+      "step": 8860
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00013134794312346992,
+      "loss": 1.4471,
+      "step": 8865
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013126218040725447,
+      "loss": 1.4217,
+      "step": 8870
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013117639219867427,
+      "loss": 1.4203,
+      "step": 8875
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013109057856768434,
+      "loss": 1.4375,
+      "step": 8880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013100473958426028,
+      "loss": 1.4119,
+      "step": 8885
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013091887531839852,
+      "loss": 1.4325,
+      "step": 8890
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013083298584011597,
+      "loss": 1.3835,
+      "step": 8895
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001307470712194502,
+      "loss": 1.4462,
+      "step": 8900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001306611315264592,
+      "loss": 1.4318,
+      "step": 8905
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013057516683122152,
+      "loss": 1.431,
+      "step": 8910
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013048917720383593,
+      "loss": 1.435,
+      "step": 8915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013040316271442173,
+      "loss": 1.4275,
+      "step": 8920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001303171234331183,
+      "loss": 1.422,
+      "step": 8925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013023105943008539,
+      "loss": 1.4462,
+      "step": 8930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001301449707755028,
+      "loss": 1.4093,
+      "step": 8935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013005885753957048,
+      "loss": 1.4153,
+      "step": 8940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012997271979250843,
+      "loss": 1.4319,
+      "step": 8945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012988655760455667,
+      "loss": 1.4326,
+      "step": 8950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000129800371045975,
+      "loss": 1.3916,
+      "step": 8955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012971416018704333,
+      "loss": 1.4428,
+      "step": 8960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012962792509806117,
+      "loss": 1.4368,
+      "step": 8965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001295416658493479,
+      "loss": 1.4149,
+      "step": 8970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001294553825112426,
+      "loss": 1.4172,
+      "step": 8975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012936907515410392,
+      "loss": 1.4337,
+      "step": 8980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012928274384831014,
+      "loss": 1.3825,
+      "step": 8985
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012919638866425913,
+      "loss": 1.4085,
+      "step": 8990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012911000967236815,
+      "loss": 1.4035,
+      "step": 8995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012902360694307387,
+      "loss": 1.4052,
+      "step": 9000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012893718054683242,
+      "loss": 1.4425,
+      "step": 9005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012885073055411903,
+      "loss": 1.4268,
+      "step": 9010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012876425703542844,
+      "loss": 1.4251,
+      "step": 9015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012867776006127428,
+      "loss": 1.4199,
+      "step": 9020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012859123970218958,
+      "loss": 1.4357,
+      "step": 9025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012850469602872623,
+      "loss": 1.4119,
+      "step": 9030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001284181291114553,
+      "loss": 1.3873,
+      "step": 9035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012833153902096664,
+      "loss": 1.3895,
+      "step": 9040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012824492582786916,
+      "loss": 1.4456,
+      "step": 9045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012815828960279047,
+      "loss": 1.41,
+      "step": 9050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012807163041637706,
+      "loss": 1.4373,
+      "step": 9055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001279849483392941,
+      "loss": 1.3842,
+      "step": 9060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012789824344222546,
+      "loss": 1.4345,
+      "step": 9065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012781151579587357,
+      "loss": 1.4377,
+      "step": 9070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012772476547095944,
+      "loss": 1.4504,
+      "step": 9075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012763799253822256,
+      "loss": 1.4384,
+      "step": 9080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012755119706842088,
+      "loss": 1.4835,
+      "step": 9085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00012746437913233066,
+      "loss": 1.4315,
+      "step": 9090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001273775388007466,
+      "loss": 1.4752,
+      "step": 9095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012729067614448156,
+      "loss": 1.4255,
+      "step": 9100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012720379123436665,
+      "loss": 1.422,
+      "step": 9105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012711688414125108,
+      "loss": 1.44,
+      "step": 9110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001270299549360022,
+      "loss": 1.4008,
+      "step": 9115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001269430036895054,
+      "loss": 1.4218,
+      "step": 9120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012685603047266398,
+      "loss": 1.4274,
+      "step": 9125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001267690353563992,
+      "loss": 1.4504,
+      "step": 9130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012668201841165017,
+      "loss": 1.4366,
+      "step": 9135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001265949797093738,
+      "loss": 1.4325,
+      "step": 9140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012650791932054473,
+      "loss": 1.4445,
+      "step": 9145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00012642083731615532,
+      "loss": 1.4038,
+      "step": 9150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001263337337672155,
+      "loss": 1.4125,
+      "step": 9155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012624660874475287,
+      "loss": 1.442,
+      "step": 9160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012615946231981238,
+      "loss": 1.4092,
+      "step": 9165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012607229456345658,
+      "loss": 1.4424,
+      "step": 9170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001259851055467653,
+      "loss": 1.4196,
+      "step": 9175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012589789534083582,
+      "loss": 1.422,
+      "step": 9180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001258106640167826,
+      "loss": 1.4399,
+      "step": 9185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001257234116457374,
+      "loss": 1.4069,
+      "step": 9190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001256361382988491,
+      "loss": 1.4012,
+      "step": 9195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012554884404728368,
+      "loss": 1.3978,
+      "step": 9200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012546152896222417,
+      "loss": 1.3993,
+      "step": 9205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012537419311487057,
+      "loss": 1.419,
+      "step": 9210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012528683657643988,
+      "loss": 1.4224,
+      "step": 9215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001251994594181659,
+      "loss": 1.4741,
+      "step": 9220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012511206171129927,
+      "loss": 1.4216,
+      "step": 9225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012502464352710742,
+      "loss": 1.4371,
+      "step": 9230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001249372049368744,
+      "loss": 1.4272,
+      "step": 9235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012484974601190097,
+      "loss": 1.4451,
+      "step": 9240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012476226682350442,
+      "loss": 1.4301,
+      "step": 9245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012467476744301866,
+      "loss": 1.4413,
+      "step": 9250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012458724794179392,
+      "loss": 1.3858,
+      "step": 9255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012449970839119697,
+      "loss": 1.4114,
+      "step": 9260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012441214886261076,
+      "loss": 1.4585,
+      "step": 9265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012432456942743477,
+      "loss": 1.4263,
+      "step": 9270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012423697015708456,
+      "loss": 1.4141,
+      "step": 9275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001241493511229918,
+      "loss": 1.4286,
+      "step": 9280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001240617123966045,
+      "loss": 1.404,
+      "step": 9285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012397405404938652,
+      "loss": 1.4444,
+      "step": 9290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012388637615281777,
+      "loss": 1.3767,
+      "step": 9295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012379867877839414,
+      "loss": 1.4326,
+      "step": 9300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012371096199762747,
+      "loss": 1.4363,
+      "step": 9305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001236232258820452,
+      "loss": 1.3958,
+      "step": 9310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001235354705031908,
+      "loss": 1.4464,
+      "step": 9315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012344769593262324,
+      "loss": 1.3948,
+      "step": 9320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001233599022419173,
+      "loss": 1.4061,
+      "step": 9325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001232720895026632,
+      "loss": 1.4299,
+      "step": 9330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012318425778646685,
+      "loss": 1.4029,
+      "step": 9335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001230964071649495,
+      "loss": 1.37,
+      "step": 9340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012300853770974787,
+      "loss": 1.3944,
+      "step": 9345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012292064949251405,
+      "loss": 1.3961,
+      "step": 9350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012283274258491543,
+      "loss": 1.3656,
+      "step": 9355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012274481705863463,
+      "loss": 1.4532,
+      "step": 9360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012265687298536942,
+      "loss": 1.4187,
+      "step": 9365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012256891043683276,
+      "loss": 1.4343,
+      "step": 9370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012248092948475263,
+      "loss": 1.4114,
+      "step": 9375
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000122392930200872,
+      "loss": 1.3942,
+      "step": 9380
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012230491265694888,
+      "loss": 1.4235,
+      "step": 9385
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000122216876924756,
+      "loss": 1.4181,
+      "step": 9390
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012212882307608116,
+      "loss": 1.4147,
+      "step": 9395
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012204075118272669,
+      "loss": 1.3878,
+      "step": 9400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001219526613165098,
+      "loss": 1.4422,
+      "step": 9405
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012186455354926228,
+      "loss": 1.451,
+      "step": 9410
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012177642795283053,
+      "loss": 1.4068,
+      "step": 9415
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012168828459907551,
+      "loss": 1.4096,
+      "step": 9420
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012160012355987265,
+      "loss": 1.4282,
+      "step": 9425
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012151194490711178,
+      "loss": 1.4695,
+      "step": 9430
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012142374871269713,
+      "loss": 1.4556,
+      "step": 9435
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012133553504854718,
+      "loss": 1.3891,
+      "step": 9440
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012124730398659474,
+      "loss": 1.4015,
+      "step": 9445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001211590555987867,
+      "loss": 1.4103,
+      "step": 9450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012107078995708417,
+      "loss": 1.4507,
+      "step": 9455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012098250713346231,
+      "loss": 1.4796,
+      "step": 9460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012089420719991022,
+      "loss": 1.4649,
+      "step": 9465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012080589022843107,
+      "loss": 1.4003,
+      "step": 9470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001207175562910418,
+      "loss": 1.4453,
+      "step": 9475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012062920545977327,
+      "loss": 1.4391,
+      "step": 9480
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012054083780667012,
+      "loss": 1.4181,
+      "step": 9485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012045245340379063,
+      "loss": 1.4388,
+      "step": 9490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001203640523232068,
+      "loss": 1.44,
+      "step": 9495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012027563463700427,
+      "loss": 1.4697,
+      "step": 9500
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012018720041728206,
+      "loss": 1.4354,
+      "step": 9505
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012009874973615287,
+      "loss": 1.3918,
+      "step": 9510
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012001028266574268,
+      "loss": 1.4441,
+      "step": 9515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011992179927819093,
+      "loss": 1.4441,
+      "step": 9520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011983329964565028,
+      "loss": 1.378,
+      "step": 9525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011974478384028672,
+      "loss": 1.4261,
+      "step": 9530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011965625193427934,
+      "loss": 1.4045,
+      "step": 9535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011956770399982045,
+      "loss": 1.4679,
+      "step": 9540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011947914010911534,
+      "loss": 1.4444,
+      "step": 9545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001193905603343824,
+      "loss": 1.434,
+      "step": 9550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011930196474785294,
+      "loss": 1.4215,
+      "step": 9555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011921335342177111,
+      "loss": 1.4442,
+      "step": 9560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011912472642839394,
+      "loss": 1.4107,
+      "step": 9565
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011903608383999125,
+      "loss": 1.3996,
+      "step": 9570
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011894742572884554,
+      "loss": 1.3999,
+      "step": 9575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011885875216725205,
+      "loss": 1.4767,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011877006322751847,
+      "loss": 1.4236,
+      "step": 9585
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011868135898196519,
+      "loss": 1.429,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011859263950292496,
+      "loss": 1.4233,
+      "step": 9595
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011850390486274303,
+      "loss": 1.4237,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011841515513377697,
+      "loss": 1.4137,
+      "step": 9605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011832639038839666,
+      "loss": 1.4343,
+      "step": 9610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00011823761069898425,
+      "loss": 1.4201,
+      "step": 9615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011814881613793404,
+      "loss": 1.4431,
+      "step": 9620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001180600067776525,
+      "loss": 1.4408,
+      "step": 9625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011797118269055812,
+      "loss": 1.4132,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001178823439490814,
+      "loss": 1.4418,
+      "step": 9635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011779349062566485,
+      "loss": 1.4081,
+      "step": 9640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011770462279276282,
+      "loss": 1.4312,
+      "step": 9645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001176157405228415,
+      "loss": 1.4477,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001175268438883789,
+      "loss": 1.4527,
+      "step": 9655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001174379329618646,
+      "loss": 1.3892,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011734900781580003,
+      "loss": 1.435,
+      "step": 9665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011726006852269804,
+      "loss": 1.4074,
+      "step": 9670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011717111515508319,
+      "loss": 1.4093,
+      "step": 9675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011708214778549131,
+      "loss": 1.4405,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011699316648646986,
+      "loss": 1.4514,
+      "step": 9685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011690417133057747,
+      "loss": 1.4149,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011681516239038423,
+      "loss": 1.4184,
+      "step": 9695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011672613973847136,
+      "loss": 1.4329,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011663710344743135,
+      "loss": 1.4197,
+      "step": 9705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011654805358986766,
+      "loss": 1.425,
+      "step": 9710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011645899023839499,
+      "loss": 1.4101,
+      "step": 9715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011636991346563893,
+      "loss": 1.4285,
+      "step": 9720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011628082334423608,
+      "loss": 1.3963,
+      "step": 9725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011619171994683389,
+      "loss": 1.4059,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011610260334609063,
+      "loss": 1.3875,
+      "step": 9735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011601347361467534,
+      "loss": 1.4455,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011592433082526781,
+      "loss": 1.4541,
+      "step": 9745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011583517505055839,
+      "loss": 1.4156,
+      "step": 9750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011574600636324813,
+      "loss": 1.3933,
+      "step": 9755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011565682483604852,
+      "loss": 1.3996,
+      "step": 9760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011556763054168154,
+      "loss": 1.4398,
+      "step": 9765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011547842355287961,
+      "loss": 1.4131,
+      "step": 9770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011538920394238551,
+      "loss": 1.3948,
+      "step": 9775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011529997178295223,
+      "loss": 1.4212,
+      "step": 9780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011521072714734309,
+      "loss": 1.4149,
+      "step": 9785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011512147010833152,
+      "loss": 1.4217,
+      "step": 9790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011503220073870111,
+      "loss": 1.4138,
+      "step": 9795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011494291911124544,
+      "loss": 1.4112,
+      "step": 9800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001148536252987682,
+      "loss": 1.3989,
+      "step": 9805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011476431937408285,
+      "loss": 1.4537,
+      "step": 9810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001146750014100129,
+      "loss": 1.4093,
+      "step": 9815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011458567147939154,
+      "loss": 1.4279,
+      "step": 9820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011449632965506183,
+      "loss": 1.4378,
+      "step": 9825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011440697600987642,
+      "loss": 1.435,
+      "step": 9830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011431761061669768,
+      "loss": 1.4388,
+      "step": 9835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011422823354839753,
+      "loss": 1.4265,
+      "step": 9840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011413884487785742,
+      "loss": 1.4337,
+      "step": 9845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011404944467796828,
+      "loss": 1.4373,
+      "step": 9850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011396003302163034,
+      "loss": 1.4282,
+      "step": 9855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011387060998175329,
+      "loss": 1.4358,
+      "step": 9860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011378117563125608,
+      "loss": 1.4105,
+      "step": 9865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011369173004306683,
+      "loss": 1.4332,
+      "step": 9870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011360227329012287,
+      "loss": 1.443,
+      "step": 9875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011351280544537064,
+      "loss": 1.4231,
+      "step": 9880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011342332658176555,
+      "loss": 1.4377,
+      "step": 9885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011333383677227214,
+      "loss": 1.4763,
+      "step": 9890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011324433608986369,
+      "loss": 1.3878,
+      "step": 9895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011315482460752252,
+      "loss": 1.4342,
+      "step": 9900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001130653023982396,
+      "loss": 1.4205,
+      "step": 9905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011297576953501481,
+      "loss": 1.4384,
+      "step": 9910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011288622609085657,
+      "loss": 1.4072,
+      "step": 9915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011279667213878205,
+      "loss": 1.417,
+      "step": 9920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011270710775181687,
+      "loss": 1.3894,
+      "step": 9925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011261753300299529,
+      "loss": 1.3755,
+      "step": 9930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011252794796535988,
+      "loss": 1.3901,
+      "step": 9935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001124383527119617,
+      "loss": 1.4307,
+      "step": 9940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011234874731586012,
+      "loss": 1.463,
+      "step": 9945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011225913185012276,
+      "loss": 1.3902,
+      "step": 9950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011216950638782545,
+      "loss": 1.3977,
+      "step": 9955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011207987100205219,
+      "loss": 1.4353,
+      "step": 9960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011199022576589506,
+      "loss": 1.4362,
+      "step": 9965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011190057075245422,
+      "loss": 1.4327,
+      "step": 9970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011181090603483768,
+      "loss": 1.4427,
+      "step": 9975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011172123168616153,
+      "loss": 1.4526,
+      "step": 9980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011163154777954956,
+      "loss": 1.3659,
+      "step": 9985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011154185438813345,
+      "loss": 1.4265,
+      "step": 9990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011145215158505258,
+      "loss": 1.4005,
+      "step": 9995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011136243944345402,
+      "loss": 1.4192,
+      "step": 10000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011127271803649243,
+      "loss": 1.4063,
+      "step": 10005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011118298743733004,
+      "loss": 1.4345,
+      "step": 10010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011109324771913659,
+      "loss": 1.4432,
+      "step": 10015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011100349895508921,
+      "loss": 1.4541,
+      "step": 10020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001109137412183725,
+      "loss": 1.4266,
+      "step": 10025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011082397458217823,
+      "loss": 1.4216,
+      "step": 10030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001107341991197056,
+      "loss": 1.4363,
+      "step": 10035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011064441490416083,
+      "loss": 1.3989,
+      "step": 10040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011055462200875743,
+      "loss": 1.4238,
+      "step": 10045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011046482050671589,
+      "loss": 1.3771,
+      "step": 10050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011037501047126379,
+      "loss": 1.4204,
+      "step": 10055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001102851919756356,
+      "loss": 1.4242,
+      "step": 10060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011019536509307276,
+      "loss": 1.443,
+      "step": 10065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011010552989682343,
+      "loss": 1.4455,
+      "step": 10070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011001568646014269,
+      "loss": 1.4177,
+      "step": 10075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010992583485629227,
+      "loss": 1.4191,
+      "step": 10080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010983597515854055,
+      "loss": 1.394,
+      "step": 10085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010974610744016254,
+      "loss": 1.4331,
+      "step": 10090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010965623177443978,
+      "loss": 1.4209,
+      "step": 10095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010956634823466028,
+      "loss": 1.4153,
+      "step": 10100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010947645689411849,
+      "loss": 1.4491,
+      "step": 10105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010938655782611517,
+      "loss": 1.3941,
+      "step": 10110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001092966511039575,
+      "loss": 1.4261,
+      "step": 10115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010920673680095874,
+      "loss": 1.4626,
+      "step": 10120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010911681499043849,
+      "loss": 1.4246,
+      "step": 10125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010902688574572233,
+      "loss": 1.4021,
+      "step": 10130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010893694914014201,
+      "loss": 1.4188,
+      "step": 10135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001088470052470352,
+      "loss": 1.4214,
+      "step": 10140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010875705413974561,
+      "loss": 1.4519,
+      "step": 10145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010866709589162276,
+      "loss": 1.4177,
+      "step": 10150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010857713057602197,
+      "loss": 1.4231,
+      "step": 10155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001084871582663044,
+      "loss": 1.4382,
+      "step": 10160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010839717903583684,
+      "loss": 1.415,
+      "step": 10165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010830719295799181,
+      "loss": 1.4295,
+      "step": 10170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00010821720010614733,
+      "loss": 1.4191,
+      "step": 10175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000108127200553687,
+      "loss": 1.3981,
+      "step": 10180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001080371943739998,
+      "loss": 1.4236,
+      "step": 10185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010794718164048026,
+      "loss": 1.4121,
+      "step": 10190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010785716242652809,
+      "loss": 1.4075,
+      "step": 10195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010776713680554842,
+      "loss": 1.4368,
+      "step": 10200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010767710485095151,
+      "loss": 1.3704,
+      "step": 10205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010758706663615284,
+      "loss": 1.4133,
+      "step": 10210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010749702223457299,
+      "loss": 1.4003,
+      "step": 10215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010740697171963754,
+      "loss": 1.4234,
+      "step": 10220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001073169151647771,
+      "loss": 1.4117,
+      "step": 10225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010722685264342722,
+      "loss": 1.4075,
+      "step": 10230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010713678422902825,
+      "loss": 1.4053,
+      "step": 10235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001070467099950254,
+      "loss": 1.4544,
+      "step": 10240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001069566300148686,
+      "loss": 1.4289,
+      "step": 10245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010686654436201249,
+      "loss": 1.4284,
+      "step": 10250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010677645310991628,
+      "loss": 1.4181,
+      "step": 10255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010668635633204384,
+      "loss": 1.4545,
+      "step": 10260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010659625410186345,
+      "loss": 1.4424,
+      "step": 10265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010650614649284791,
+      "loss": 1.4275,
+      "step": 10270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010641603357847434,
+      "loss": 1.4008,
+      "step": 10275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010632591543222426,
+      "loss": 1.4401,
+      "step": 10280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010623579212758336,
+      "loss": 1.3828,
+      "step": 10285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010614566373804167,
+      "loss": 1.4171,
+      "step": 10290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010605553033709321,
+      "loss": 1.4277,
+      "step": 10295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001059653919982362,
+      "loss": 1.4299,
+      "step": 10300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010587524879497286,
+      "loss": 1.4266,
+      "step": 10305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010578510080080937,
+      "loss": 1.418,
+      "step": 10310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001056949480892558,
+      "loss": 1.4423,
+      "step": 10315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010560479073382605,
+      "loss": 1.4533,
+      "step": 10320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010551462880803793,
+      "loss": 1.4045,
+      "step": 10325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001054244623854128,
+      "loss": 1.4342,
+      "step": 10330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010533429153947582,
+      "loss": 1.4028,
+      "step": 10335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010524411634375566,
+      "loss": 1.4093,
+      "step": 10340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010515393687178467,
+      "loss": 1.4353,
+      "step": 10345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010506375319709852,
+      "loss": 1.4377,
+      "step": 10350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010497356539323643,
+      "loss": 1.4349,
+      "step": 10355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010488337353374093,
+      "loss": 1.421,
+      "step": 10360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010479317769215793,
+      "loss": 1.4066,
+      "step": 10365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010470297794203643,
+      "loss": 1.4299,
+      "step": 10370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010461277435692882,
+      "loss": 1.4323,
+      "step": 10375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010452256701039045,
+      "loss": 1.4183,
+      "step": 10380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010443235597597985,
+      "loss": 1.4495,
+      "step": 10385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010434214132725846,
+      "loss": 1.4152,
+      "step": 10390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010425192313779075,
+      "loss": 1.4081,
+      "step": 10395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010416170148114404,
+      "loss": 1.4235,
+      "step": 10400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001040714764308885,
+      "loss": 1.3903,
+      "step": 10405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010398124806059701,
+      "loss": 1.4415,
+      "step": 10410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010389101644384524,
+      "loss": 1.4446,
+      "step": 10415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010380078165421144,
+      "loss": 1.4458,
+      "step": 10420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010371054376527647,
+      "loss": 1.4719,
+      "step": 10425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010362030285062369,
+      "loss": 1.4487,
+      "step": 10430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010353005898383905,
+      "loss": 1.4335,
+      "step": 10435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010343981223851074,
+      "loss": 1.4065,
+      "step": 10440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010334956268822937,
+      "loss": 1.4032,
+      "step": 10445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010325931040658783,
+      "loss": 1.4184,
+      "step": 10450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010316905546718128,
+      "loss": 1.4337,
+      "step": 10455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010307879794360701,
+      "loss": 1.3685,
+      "step": 10460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001029885379094644,
+      "loss": 1.4441,
+      "step": 10465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010289827543835493,
+      "loss": 1.4085,
+      "step": 10470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010280801060388199,
+      "loss": 1.4465,
+      "step": 10475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010271774347965097,
+      "loss": 1.4511,
+      "step": 10480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010262747413926907,
+      "loss": 1.4433,
+      "step": 10485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010253720265634537,
+      "loss": 1.4181,
+      "step": 10490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010244692910449061,
+      "loss": 1.4608,
+      "step": 10495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00010235665355731727,
+      "loss": 1.4125,
+      "step": 10500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010226637608843947,
+      "loss": 1.4069,
+      "step": 10505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010217609677147287,
+      "loss": 1.4075,
+      "step": 10510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010208581568003459,
+      "loss": 1.3924,
+      "step": 10515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010199553288774333,
+      "loss": 1.4404,
+      "step": 10520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010190524846821903,
+      "loss": 1.4055,
+      "step": 10525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010181496249508305,
+      "loss": 1.4068,
+      "step": 10530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010172467504195798,
+      "loss": 1.4726,
+      "step": 10535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010163438618246763,
+      "loss": 1.4363,
+      "step": 10540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010154409599023693,
+      "loss": 1.4466,
+      "step": 10545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010145380453889195,
+      "loss": 1.4278,
+      "step": 10550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010136351190205975,
+      "loss": 1.4285,
+      "step": 10555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010127321815336837,
+      "loss": 1.4437,
+      "step": 10560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010118292336644668,
+      "loss": 1.3674,
+      "step": 10565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010109262761492458,
+      "loss": 1.3475,
+      "step": 10570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010100233097243255,
+      "loss": 1.4186,
+      "step": 10575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010091203351260194,
+      "loss": 1.3973,
+      "step": 10580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00010082173530906467,
+      "loss": 1.4284,
+      "step": 10585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010073143643545339,
+      "loss": 1.4689,
+      "step": 10590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010064113696540111,
+      "loss": 1.3903,
+      "step": 10595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010055083697254156,
+      "loss": 1.4076,
+      "step": 10600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00010046053653050862,
+      "loss": 1.4381,
+      "step": 10605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010037023571293682,
+      "loss": 1.4145,
+      "step": 10610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010027993459346079,
+      "loss": 1.4353,
+      "step": 10615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010018963324571551,
+      "loss": 1.4071,
+      "step": 10620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010009933174333608,
+      "loss": 1.3875,
+      "step": 10625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010000903015995783,
+      "loss": 1.4194,
+      "step": 10630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.991872856921601e-05,
+      "loss": 1.4326,
+      "step": 10635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.9828427044746e-05,
+      "loss": 1.4118,
+      "step": 10640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.973812566018309e-05,
+      "loss": 1.4773,
+      "step": 10645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.96478244891624e-05,
+      "loss": 1.4413,
+      "step": 10650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.955752360531896e-05,
+      "loss": 1.3984,
+      "step": 10655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.94672230822875e-05,
+      "loss": 1.4695,
+      "step": 10660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.937692299370251e-05,
+      "loss": 1.404,
+      "step": 10665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.928662341319808e-05,
+      "loss": 1.4331,
+      "step": 10670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.919632441440791e-05,
+      "loss": 1.441,
+      "step": 10675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.910602607096522e-05,
+      "loss": 1.4424,
+      "step": 10680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.90157284565027e-05,
+      "loss": 1.4336,
+      "step": 10685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.892543164465243e-05,
+      "loss": 1.4476,
+      "step": 10690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.883513570904587e-05,
+      "loss": 1.4135,
+      "step": 10695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5,
+      "learning_rate": 9.874484072331371e-05,
+      "loss": 1.3951,
+      "step": 10700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.865454676108592e-05,
+      "loss": 1.4369,
+      "step": 10705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.856425389599159e-05,
+      "loss": 1.3727,
+      "step": 10710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.847396220165898e-05,
+      "loss": 1.4404,
+      "step": 10715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.838367175171531e-05,
+      "loss": 1.4034,
+      "step": 10720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.829338261978686e-05,
+      "loss": 1.4248,
+      "step": 10725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.82030948794988e-05,
+      "loss": 1.4239,
+      "step": 10730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.811280860447515e-05,
+      "loss": 1.3853,
+      "step": 10735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.802252386833875e-05,
+      "loss": 1.394,
+      "step": 10740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.793224074471125e-05,
+      "loss": 1.4399,
+      "step": 10745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.784195930721284e-05,
+      "loss": 1.3797,
+      "step": 10750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.775167962946248e-05,
+      "loss": 1.3967,
+      "step": 10755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.76614017850776e-05,
+      "loss": 1.4251,
+      "step": 10760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.757112584767422e-05,
+      "loss": 1.3978,
+      "step": 10765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.748085189086668e-05,
+      "loss": 1.4219,
+      "step": 10770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.739057998826786e-05,
+      "loss": 1.41,
+      "step": 10775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.730031021348881e-05,
+      "loss": 1.4403,
+      "step": 10780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.721004264013899e-05,
+      "loss": 1.3703,
+      "step": 10785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.711977734182593e-05,
+      "loss": 1.4361,
+      "step": 10790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.702951439215543e-05,
+      "loss": 1.4309,
+      "step": 10795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.693925386473127e-05,
+      "loss": 1.3903,
+      "step": 10800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.684899583315531e-05,
+      "loss": 1.4016,
+      "step": 10805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.67587403710274e-05,
+      "loss": 1.4281,
+      "step": 10810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.666848755194519e-05,
+      "loss": 1.4201,
+      "step": 10815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.65782374495043e-05,
+      "loss": 1.4347,
+      "step": 10820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.648799013729802e-05,
+      "loss": 1.4321,
+      "step": 10825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.63977456889175e-05,
+      "loss": 1.4376,
+      "step": 10830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.630750417795141e-05,
+      "loss": 1.4297,
+      "step": 10835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.621726567798614e-05,
+      "loss": 1.457,
+      "step": 10840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.612703026260553e-05,
+      "loss": 1.3704,
+      "step": 10845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.603679800539102e-05,
+      "loss": 1.3975,
+      "step": 10850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.594656897992133e-05,
+      "loss": 1.4303,
+      "step": 10855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.585634325977268e-05,
+      "loss": 1.4059,
+      "step": 10860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.57661209185185e-05,
+      "loss": 1.4084,
+      "step": 10865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.567590202972952e-05,
+      "loss": 1.4019,
+      "step": 10870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.558568666697362e-05,
+      "loss": 1.4165,
+      "step": 10875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.549547490381585e-05,
+      "loss": 1.4164,
+      "step": 10880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.540526681381824e-05,
+      "loss": 1.3971,
+      "step": 10885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.531506247053995e-05,
+      "loss": 1.4246,
+      "step": 10890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.522486194753695e-05,
+      "loss": 1.4365,
+      "step": 10895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.513466531836221e-05,
+      "loss": 1.3987,
+      "step": 10900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.504447265656544e-05,
+      "loss": 1.4129,
+      "step": 10905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.495428403569317e-05,
+      "loss": 1.4259,
+      "step": 10910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.486409952928858e-05,
+      "loss": 1.429,
+      "step": 10915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.477391921089158e-05,
+      "loss": 1.4268,
+      "step": 10920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.468374315403858e-05,
+      "loss": 1.3592,
+      "step": 10925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.459357143226255e-05,
+      "loss": 1.4077,
+      "step": 10930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.450340411909293e-05,
+      "loss": 1.4344,
+      "step": 10935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.441324128805555e-05,
+      "loss": 1.3846,
+      "step": 10940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.432308301267261e-05,
+      "loss": 1.3887,
+      "step": 10945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.423292936646257e-05,
+      "loss": 1.4068,
+      "step": 10950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.414278042294012e-05,
+      "loss": 1.389,
+      "step": 10955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.405263625561613e-05,
+      "loss": 1.408,
+      "step": 10960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.396249693799754e-05,
+      "loss": 1.4058,
+      "step": 10965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.387236254358741e-05,
+      "loss": 1.4122,
+      "step": 10970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.378223314588467e-05,
+      "loss": 1.3976,
+      "step": 10975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.36921088183843e-05,
+      "loss": 1.439,
+      "step": 10980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.360198963457705e-05,
+      "loss": 1.4165,
+      "step": 10985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.351187566794953e-05,
+      "loss": 1.4397,
+      "step": 10990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.342176699198406e-05,
+      "loss": 1.4671,
+      "step": 10995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.333166368015869e-05,
+      "loss": 1.44,
+      "step": 11000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.324156580594704e-05,
+      "loss": 1.4439,
+      "step": 11005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.315147344281836e-05,
+      "loss": 1.4235,
+      "step": 11010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.306138666423733e-05,
+      "loss": 1.4024,
+      "step": 11015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.297130554366413e-05,
+      "loss": 1.3865,
+      "step": 11020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.288123015455436e-05,
+      "loss": 1.4025,
+      "step": 11025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.279116057035882e-05,
+      "loss": 1.4177,
+      "step": 11030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.270109686452375e-05,
+      "loss": 1.4183,
+      "step": 11035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.261103911049041e-05,
+      "loss": 1.4063,
+      "step": 11040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.252098738169538e-05,
+      "loss": 1.4324,
+      "step": 11045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.24309417515702e-05,
+      "loss": 1.469,
+      "step": 11050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.234090229354149e-05,
+      "loss": 1.3761,
+      "step": 11055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.225086908103082e-05,
+      "loss": 1.386,
+      "step": 11060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.216084218745472e-05,
+      "loss": 1.4339,
+      "step": 11065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.207082168622448e-05,
+      "loss": 1.3859,
+      "step": 11070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.198080765074625e-05,
+      "loss": 1.396,
+      "step": 11075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.189080015442085e-05,
+      "loss": 1.3841,
+      "step": 11080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.180079927064386e-05,
+      "loss": 1.382,
+      "step": 11085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.171080507280532e-05,
+      "loss": 1.4463,
+      "step": 11090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.162081763428999e-05,
+      "loss": 1.4456,
+      "step": 11095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.153083702847695e-05,
+      "loss": 1.4228,
+      "step": 11100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.14408633287399e-05,
+      "loss": 1.3921,
+      "step": 11105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.135089660844669e-05,
+      "loss": 1.4088,
+      "step": 11110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.126093694095961e-05,
+      "loss": 1.3919,
+      "step": 11115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.117098439963522e-05,
+      "loss": 1.4446,
+      "step": 11120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.108103905782419e-05,
+      "loss": 1.3774,
+      "step": 11125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.099110098887136e-05,
+      "loss": 1.407,
+      "step": 11130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.090117026611564e-05,
+      "loss": 1.4042,
+      "step": 11135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.081124696288995e-05,
+      "loss": 1.4179,
+      "step": 11140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.072133115252112e-05,
+      "loss": 1.4189,
+      "step": 11145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.063142290832997e-05,
+      "loss": 1.3894,
+      "step": 11150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.054152230363102e-05,
+      "loss": 1.4122,
+      "step": 11155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.045162941173266e-05,
+      "loss": 1.406,
+      "step": 11160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.036174430593694e-05,
+      "loss": 1.4228,
+      "step": 11165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.027186705953958e-05,
+      "loss": 1.4127,
+      "step": 11170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.018199774582988e-05,
+      "loss": 1.3923,
+      "step": 11175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.009213643809072e-05,
+      "loss": 1.4126,
+      "step": 11180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.000228320959833e-05,
+      "loss": 1.3914,
+      "step": 11185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.991243813362252e-05,
+      "loss": 1.4298,
+      "step": 11190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.982260128342628e-05,
+      "loss": 1.42,
+      "step": 11195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.973277273226607e-05,
+      "loss": 1.4267,
+      "step": 11200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.96429525533914e-05,
+      "loss": 1.4044,
+      "step": 11205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.95531408200451e-05,
+      "loss": 1.3861,
+      "step": 11210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.946333760546303e-05,
+      "loss": 1.4079,
+      "step": 11215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.937354298287414e-05,
+      "loss": 1.4124,
+      "step": 11220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.928375702550036e-05,
+      "loss": 1.3697,
+      "step": 11225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.919397980655657e-05,
+      "loss": 1.4066,
+      "step": 11230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.910421139925045e-05,
+      "loss": 1.4038,
+      "step": 11235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.901445187678264e-05,
+      "loss": 1.4115,
+      "step": 11240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.892470131234639e-05,
+      "loss": 1.4057,
+      "step": 11245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.883495977912775e-05,
+      "loss": 1.4335,
+      "step": 11250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.874522735030532e-05,
+      "loss": 1.3702,
+      "step": 11255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.865550409905037e-05,
+      "loss": 1.3975,
+      "step": 11260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.856579009852657e-05,
+      "loss": 1.4373,
+      "step": 11265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.847608542189017e-05,
+      "loss": 1.418,
+      "step": 11270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.83863901422897e-05,
+      "loss": 1.4033,
+      "step": 11275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.829670433286613e-05,
+      "loss": 1.4357,
+      "step": 11280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.820702806675263e-05,
+      "loss": 1.4209,
+      "step": 11285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.811736141707466e-05,
+      "loss": 1.4346,
+      "step": 11290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.802770445694975e-05,
+      "loss": 1.4062,
+      "step": 11295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.793805725948764e-05,
+      "loss": 1.4005,
+      "step": 11300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.784841989778996e-05,
+      "loss": 1.4151,
+      "step": 11305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.775879244495052e-05,
+      "loss": 1.4449,
+      "step": 11310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.766917497405481e-05,
+      "loss": 1.4397,
+      "step": 11315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.757956755818041e-05,
+      "loss": 1.424,
+      "step": 11320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.748997027039653e-05,
+      "loss": 1.4169,
+      "step": 11325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.740038318376423e-05,
+      "loss": 1.4071,
+      "step": 11330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.731080637133618e-05,
+      "loss": 1.4427,
+      "step": 11335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.722123990615673e-05,
+      "loss": 1.4193,
+      "step": 11340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.713168386126173e-05,
+      "loss": 1.3958,
+      "step": 11345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.704213830967861e-05,
+      "loss": 1.457,
+      "step": 11350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.695260332442616e-05,
+      "loss": 1.4133,
+      "step": 11355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.686307897851463e-05,
+      "loss": 1.4035,
+      "step": 11360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.677356534494553e-05,
+      "loss": 1.3896,
+      "step": 11365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.668406249671169e-05,
+      "loss": 1.4115,
+      "step": 11370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.65945705067971e-05,
+      "loss": 1.4179,
+      "step": 11375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.650508944817692e-05,
+      "loss": 1.4196,
+      "step": 11380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.641561939381737e-05,
+      "loss": 1.4511,
+      "step": 11385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.632616041667577e-05,
+      "loss": 1.4401,
+      "step": 11390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.623671258970028e-05,
+      "loss": 1.4221,
+      "step": 11395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.614727598583015e-05,
+      "loss": 1.4375,
+      "step": 11400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.605785067799527e-05,
+      "loss": 1.3985,
+      "step": 11405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.596843673911643e-05,
+      "loss": 1.4173,
+      "step": 11410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.58790342421052e-05,
+      "loss": 1.443,
+      "step": 11415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.578964325986368e-05,
+      "loss": 1.3976,
+      "step": 11420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.570026386528475e-05,
+      "loss": 1.3657,
+      "step": 11425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.561089613125166e-05,
+      "loss": 1.4348,
+      "step": 11430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.55215401306383e-05,
+      "loss": 1.4046,
+      "step": 11435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.543219593630892e-05,
+      "loss": 1.4324,
+      "step": 11440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.534286362111812e-05,
+      "loss": 1.42,
+      "step": 11445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.525354325791092e-05,
+      "loss": 1.4366,
+      "step": 11450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.516423491952247e-05,
+      "loss": 1.4362,
+      "step": 11455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.50749386787782e-05,
+      "loss": 1.384,
+      "step": 11460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.498565460849362e-05,
+      "loss": 1.3842,
+      "step": 11465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.489638278147433e-05,
+      "loss": 1.4209,
+      "step": 11470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.480712327051599e-05,
+      "loss": 1.3873,
+      "step": 11475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.471787614840416e-05,
+      "loss": 1.4494,
+      "step": 11480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.462864148791432e-05,
+      "loss": 1.422,
+      "step": 11485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.453941936181181e-05,
+      "loss": 1.434,
+      "step": 11490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.445020984285169e-05,
+      "loss": 1.4391,
+      "step": 11495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.436101300377881e-05,
+      "loss": 1.3965,
+      "step": 11500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.427182891732762e-05,
+      "loss": 1.4083,
+      "step": 11505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.418265765622225e-05,
+      "loss": 1.3746,
+      "step": 11510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.409349929317623e-05,
+      "loss": 1.3924,
+      "step": 11515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.400435390089277e-05,
+      "loss": 1.3971,
+      "step": 11520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.391522155206429e-05,
+      "loss": 1.395,
+      "step": 11525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.382610231937276e-05,
+      "loss": 1.439,
+      "step": 11530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.373699627548934e-05,
+      "loss": 1.425,
+      "step": 11535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.364790349307448e-05,
+      "loss": 1.389,
+      "step": 11540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.355882404477778e-05,
+      "loss": 1.4014,
+      "step": 11545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.346975800323804e-05,
+      "loss": 1.4194,
+      "step": 11550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.338070544108304e-05,
+      "loss": 1.4122,
+      "step": 11555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.329166643092963e-05,
+      "loss": 1.4194,
+      "step": 11560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.320264104538357e-05,
+      "loss": 1.3873,
+      "step": 11565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.311362935703955e-05,
+      "loss": 1.4282,
+      "step": 11570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.302463143848102e-05,
+      "loss": 1.4116,
+      "step": 11575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.293564736228034e-05,
+      "loss": 1.4194,
+      "step": 11580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.284667720099839e-05,
+      "loss": 1.4211,
+      "step": 11585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.275772102718489e-05,
+      "loss": 1.4032,
+      "step": 11590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.2668778913378e-05,
+      "loss": 1.3712,
+      "step": 11595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.257985093210455e-05,
+      "loss": 1.4273,
+      "step": 11600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.249093715587972e-05,
+      "loss": 1.4499,
+      "step": 11605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.240203765720722e-05,
+      "loss": 1.3837,
+      "step": 11610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.231315250857902e-05,
+      "loss": 1.4142,
+      "step": 11615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.222428178247548e-05,
+      "loss": 1.3888,
+      "step": 11620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.21354255513651e-05,
+      "loss": 1.384,
+      "step": 11625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.204658388770466e-05,
+      "loss": 1.3775,
+      "step": 11630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.195775686393897e-05,
+      "loss": 1.4199,
+      "step": 11635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.1868944552501e-05,
+      "loss": 1.3916,
+      "step": 11640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.178014702581162e-05,
+      "loss": 1.3924,
+      "step": 11645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.169136435627971e-05,
+      "loss": 1.4131,
+      "step": 11650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.160259661630201e-05,
+      "loss": 1.3814,
+      "step": 11655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.151384387826313e-05,
+      "loss": 1.3941,
+      "step": 11660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.142510621453536e-05,
+      "loss": 1.395,
+      "step": 11665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.13363836974788e-05,
+      "loss": 1.3861,
+      "step": 11670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.124767639944109e-05,
+      "loss": 1.4077,
+      "step": 11675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.115898439275756e-05,
+      "loss": 1.4453,
+      "step": 11680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5,
+      "learning_rate": 8.107030774975101e-05,
+      "loss": 1.4159,
+      "step": 11685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.098164654273174e-05,
+      "loss": 1.4389,
+      "step": 11690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.089300084399747e-05,
+      "loss": 1.3747,
+      "step": 11695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.08043707258332e-05,
+      "loss": 1.4058,
+      "step": 11700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.071575626051133e-05,
+      "loss": 1.4204,
+      "step": 11705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.062715752029142e-05,
+      "loss": 1.4552,
+      "step": 11710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.053857457742025e-05,
+      "loss": 1.4289,
+      "step": 11715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.045000750413169e-05,
+      "loss": 1.4027,
+      "step": 11720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.036145637264673e-05,
+      "loss": 1.4515,
+      "step": 11725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.027292125517324e-05,
+      "loss": 1.4288,
+      "step": 11730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.018440222390616e-05,
+      "loss": 1.4238,
+      "step": 11735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.009589935102723e-05,
+      "loss": 1.4264,
+      "step": 11740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.000741270870507e-05,
+      "loss": 1.4174,
+      "step": 11745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.991894236909498e-05,
+      "loss": 1.4294,
+      "step": 11750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.98304884043391e-05,
+      "loss": 1.4169,
+      "step": 11755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.974205088656606e-05,
+      "loss": 1.4359,
+      "step": 11760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.965362988789121e-05,
+      "loss": 1.4157,
+      "step": 11765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.956522548041635e-05,
+      "loss": 1.4213,
+      "step": 11770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.947683773622982e-05,
+      "loss": 1.4131,
+      "step": 11775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.938846672740627e-05,
+      "loss": 1.4234,
+      "step": 11780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.930011252600683e-05,
+      "loss": 1.4135,
+      "step": 11785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.92117752040788e-05,
+      "loss": 1.4611,
+      "step": 11790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.912345483365581e-05,
+      "loss": 1.4269,
+      "step": 11795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.903515148675762e-05,
+      "loss": 1.4278,
+      "step": 11800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.894686523539013e-05,
+      "loss": 1.3783,
+      "step": 11805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.885859615154527e-05,
+      "loss": 1.4143,
+      "step": 11810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.877034430720102e-05,
+      "loss": 1.4065,
+      "step": 11815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5,
+      "learning_rate": 7.868210977432123e-05,
+      "loss": 1.389,
+      "step": 11820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.85938926248557e-05,
+      "loss": 1.4302,
+      "step": 11825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.850569293074006e-05,
+      "loss": 1.416,
+      "step": 11830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.841751076389563e-05,
+      "loss": 1.3923,
+      "step": 11835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.832934619622954e-05,
+      "loss": 1.4171,
+      "step": 11840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.824119929963444e-05,
+      "loss": 1.4211,
+      "step": 11845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.81530701459887e-05,
+      "loss": 1.4153,
+      "step": 11850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.806495880715614e-05,
+      "loss": 1.4169,
+      "step": 11855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.797686535498611e-05,
+      "loss": 1.4031,
+      "step": 11860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.788878986131331e-05,
+      "loss": 1.4202,
+      "step": 11865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.780073239795787e-05,
+      "loss": 1.3808,
+      "step": 11870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.771269303672513e-05,
+      "loss": 1.4002,
+      "step": 11875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.762467184940574e-05,
+      "loss": 1.3846,
+      "step": 11880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.75366689077755e-05,
+      "loss": 1.4054,
+      "step": 11885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.744868428359536e-05,
+      "loss": 1.3909,
+      "step": 11890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.736071804861127e-05,
+      "loss": 1.4172,
+      "step": 11895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.727277027455428e-05,
+      "loss": 1.3996,
+      "step": 11900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.718484103314026e-05,
+      "loss": 1.4351,
+      "step": 11905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.709693039607012e-05,
+      "loss": 1.3794,
+      "step": 11910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.700903843502947e-05,
+      "loss": 1.3866,
+      "step": 11915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.692116522168877e-05,
+      "loss": 1.428,
+      "step": 11920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.683331082770311e-05,
+      "loss": 1.4656,
+      "step": 11925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.674547532471235e-05,
+      "loss": 1.4387,
+      "step": 11930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.665765878434084e-05,
+      "loss": 1.4312,
+      "step": 11935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.656986127819754e-05,
+      "loss": 1.4365,
+      "step": 11940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.648208287787584e-05,
+      "loss": 1.3558,
+      "step": 11945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.639432365495357e-05,
+      "loss": 1.3889,
+      "step": 11950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.630658368099291e-05,
+      "loss": 1.422,
+      "step": 11955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.62188630275404e-05,
+      "loss": 1.3943,
+      "step": 11960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.613116176612672e-05,
+      "loss": 1.3815,
+      "step": 11965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.604347996826682e-05,
+      "loss": 1.4084,
+      "step": 11970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.595581770545978e-05,
+      "loss": 1.4034,
+      "step": 11975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.58681750491887e-05,
+      "loss": 1.412,
+      "step": 11980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.578055207092071e-05,
+      "loss": 1.3972,
+      "step": 11985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.569294884210694e-05,
+      "loss": 1.3899,
+      "step": 11990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.560536543418235e-05,
+      "loss": 1.4266,
+      "step": 11995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.551780191856575e-05,
+      "loss": 1.4088,
+      "step": 12000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.543025836665977e-05,
+      "loss": 1.4214,
+      "step": 12005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.53427348498507e-05,
+      "loss": 1.4133,
+      "step": 12010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.525523143950859e-05,
+      "loss": 1.3915,
+      "step": 12015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.516774820698695e-05,
+      "loss": 1.4307,
+      "step": 12020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.5080285223623e-05,
+      "loss": 1.4384,
+      "step": 12025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.499284256073731e-05,
+      "loss": 1.3982,
+      "step": 12030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.490542028963396e-05,
+      "loss": 1.4055,
+      "step": 12035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.481801848160035e-05,
+      "loss": 1.4069,
+      "step": 12040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.473063720790727e-05,
+      "loss": 1.3999,
+      "step": 12045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.464327653980865e-05,
+      "loss": 1.4145,
+      "step": 12050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.455593654854176e-05,
+      "loss": 1.4143,
+      "step": 12055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.446861730532688e-05,
+      "loss": 1.435,
+      "step": 12060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.438131888136746e-05,
+      "loss": 1.4205,
+      "step": 12065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.429404134784987e-05,
+      "loss": 1.389,
+      "step": 12070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.420678477594361e-05,
+      "loss": 1.4084,
+      "step": 12075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.411954923680091e-05,
+      "loss": 1.4572,
+      "step": 12080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.403233480155697e-05,
+      "loss": 1.3838,
+      "step": 12085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5,
+      "learning_rate": 7.394514154132975e-05,
+      "loss": 1.3921,
+      "step": 12090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.385796952721991e-05,
+      "loss": 1.4163,
+      "step": 12095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.377081883031079e-05,
+      "loss": 1.4066,
+      "step": 12100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.368368952166839e-05,
+      "loss": 1.43,
+      "step": 12105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.359658167234125e-05,
+      "loss": 1.4091,
+      "step": 12110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.350949535336041e-05,
+      "loss": 1.3923,
+      "step": 12115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.342243063573932e-05,
+      "loss": 1.4047,
+      "step": 12120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.333538759047389e-05,
+      "loss": 1.3858,
+      "step": 12125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.324836628854226e-05,
+      "loss": 1.407,
+      "step": 12130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.316136680090494e-05,
+      "loss": 1.4015,
+      "step": 12135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.307438919850456e-05,
+      "loss": 1.3805,
+      "step": 12140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.298743355226599e-05,
+      "loss": 1.4403,
+      "step": 12145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.290049993309611e-05,
+      "loss": 1.4015,
+      "step": 12150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.281358841188392e-05,
+      "loss": 1.4074,
+      "step": 12155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.272669905950036e-05,
+      "loss": 1.409,
+      "step": 12160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.263983194679827e-05,
+      "loss": 1.4018,
+      "step": 12165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.25529871446124e-05,
+      "loss": 1.4238,
+      "step": 12170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.246616472375928e-05,
+      "loss": 1.4122,
+      "step": 12175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.237936475503719e-05,
+      "loss": 1.4233,
+      "step": 12180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.229258730922615e-05,
+      "loss": 1.4236,
+      "step": 12185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.22058324570877e-05,
+      "loss": 1.3993,
+      "step": 12190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.21191002693651e-05,
+      "loss": 1.4025,
+      "step": 12195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.203239081678299e-05,
+      "loss": 1.4303,
+      "step": 12200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.194570417004759e-05,
+      "loss": 1.3899,
+      "step": 12205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.185904039984648e-05,
+      "loss": 1.4111,
+      "step": 12210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.177239957684851e-05,
+      "loss": 1.3872,
+      "step": 12215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.168578177170397e-05,
+      "loss": 1.3984,
+      "step": 12220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.159918705504424e-05,
+      "loss": 1.4333,
+      "step": 12225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.151261549748195e-05,
+      "loss": 1.4313,
+      "step": 12230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.14260671696108e-05,
+      "loss": 1.4067,
+      "step": 12235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.13395421420056e-05,
+      "loss": 1.4108,
+      "step": 12240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.125304048522211e-05,
+      "loss": 1.4413,
+      "step": 12245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.116656226979708e-05,
+      "loss": 1.3683,
+      "step": 12250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.108010756624808e-05,
+      "loss": 1.3989,
+      "step": 12255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.099367644507357e-05,
+      "loss": 1.4072,
+      "step": 12260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.090726897675277e-05,
+      "loss": 1.3872,
+      "step": 12265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.082088523174558e-05,
+      "loss": 1.4026,
+      "step": 12270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.073452528049254e-05,
+      "loss": 1.3791,
+      "step": 12275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.06481891934149e-05,
+      "loss": 1.3941,
+      "step": 12280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.056187704091434e-05,
+      "loss": 1.3936,
+      "step": 12285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.047558889337302e-05,
+      "loss": 1.4152,
+      "step": 12290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.03893248211536e-05,
+      "loss": 1.4064,
+      "step": 12295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.030308489459904e-05,
+      "loss": 1.394,
+      "step": 12300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.021686918403266e-05,
+      "loss": 1.4682,
+      "step": 12305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.013067775975799e-05,
+      "loss": 1.415,
+      "step": 12310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.004451069205881e-05,
+      "loss": 1.3852,
+      "step": 12315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.995836805119897e-05,
+      "loss": 1.4032,
+      "step": 12320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.98722499074225e-05,
+      "loss": 1.4187,
+      "step": 12325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.978615633095331e-05,
+      "loss": 1.4065,
+      "step": 12330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.970008739199543e-05,
+      "loss": 1.418,
+      "step": 12335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.961404316073267e-05,
+      "loss": 1.4124,
+      "step": 12340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.95280237073288e-05,
+      "loss": 1.4179,
+      "step": 12345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.944202910192732e-05,
+      "loss": 1.4075,
+      "step": 12350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.93560594146515e-05,
+      "loss": 1.367,
+      "step": 12355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.927011471560422e-05,
+      "loss": 1.4334,
+      "step": 12360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.918419507486813e-05,
+      "loss": 1.4224,
+      "step": 12365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 1.3795,
+      "step": 12370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.901243124855733e-05,
+      "loss": 1.3588,
+      "step": 12375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.892658720304535e-05,
+      "loss": 1.4085,
+      "step": 12380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.884076849596988e-05,
+      "loss": 1.3963,
+      "step": 12385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.875497519731067e-05,
+      "loss": 1.3857,
+      "step": 12390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.866920737702688e-05,
+      "loss": 1.4228,
+      "step": 12395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.858346510505678e-05,
+      "loss": 1.3961,
+      "step": 12400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.849774845131791e-05,
+      "loss": 1.425,
+      "step": 12405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.841205748570685e-05,
+      "loss": 1.3729,
+      "step": 12410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.832639227809927e-05,
+      "loss": 1.4257,
+      "step": 12415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.82407528983498e-05,
+      "loss": 1.3877,
+      "step": 12420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.815513941629204e-05,
+      "loss": 1.4204,
+      "step": 12425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.806955190173848e-05,
+      "loss": 1.431,
+      "step": 12430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.798399042448039e-05,
+      "loss": 1.3797,
+      "step": 12435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.789845505428782e-05,
+      "loss": 1.4246,
+      "step": 12440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.781294586090962e-05,
+      "loss": 1.3974,
+      "step": 12445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.772746291407315e-05,
+      "loss": 1.4657,
+      "step": 12450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.764200628348449e-05,
+      "loss": 1.3938,
+      "step": 12455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.755657603882816e-05,
+      "loss": 1.3811,
+      "step": 12460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.747117224976726e-05,
+      "loss": 1.3983,
+      "step": 12465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.738579498594322e-05,
+      "loss": 1.4247,
+      "step": 12470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.730044431697595e-05,
+      "loss": 1.3805,
+      "step": 12475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.721512031246358e-05,
+      "loss": 1.395,
+      "step": 12480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.712982304198254e-05,
+      "loss": 1.417,
+      "step": 12485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.704455257508743e-05,
+      "loss": 1.4102,
+      "step": 12490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.695930898131107e-05,
+      "loss": 1.4234,
+      "step": 12495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.687409233016422e-05,
+      "loss": 1.4232,
+      "step": 12500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.678890269113587e-05,
+      "loss": 1.396,
+      "step": 12505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.670374013369279e-05,
+      "loss": 1.4411,
+      "step": 12510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.661860472727981e-05,
+      "loss": 1.4217,
+      "step": 12515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.65334965413195e-05,
+      "loss": 1.4315,
+      "step": 12520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.644841564521237e-05,
+      "loss": 1.4366,
+      "step": 12525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.636336210833654e-05,
+      "loss": 1.4163,
+      "step": 12530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.627833600004791e-05,
+      "loss": 1.4071,
+      "step": 12535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.619333738967996e-05,
+      "loss": 1.3856,
+      "step": 12540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.610836634654382e-05,
+      "loss": 1.441,
+      "step": 12545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.602342293992805e-05,
+      "loss": 1.409,
+      "step": 12550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.593850723909875e-05,
+      "loss": 1.3984,
+      "step": 12555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.585361931329937e-05,
+      "loss": 1.3951,
+      "step": 12560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.576875923175075e-05,
+      "loss": 1.4047,
+      "step": 12565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.568392706365099e-05,
+      "loss": 1.3699,
+      "step": 12570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.559912287817547e-05,
+      "loss": 1.3845,
+      "step": 12575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.551434674447676e-05,
+      "loss": 1.3977,
+      "step": 12580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.542959873168446e-05,
+      "loss": 1.4172,
+      "step": 12585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.534487890890536e-05,
+      "loss": 1.4026,
+      "step": 12590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.526018734522317e-05,
+      "loss": 1.4244,
+      "step": 12595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.517552410969863e-05,
+      "loss": 1.4092,
+      "step": 12600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.50908892713693e-05,
+      "loss": 1.4277,
+      "step": 12605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.50062828992497e-05,
+      "loss": 1.4094,
+      "step": 12610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.4921705062331e-05,
+      "loss": 1.4182,
+      "step": 12615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.48371558295812e-05,
+      "loss": 1.4341,
+      "step": 12620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.475263526994494e-05,
+      "loss": 1.3804,
+      "step": 12625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.466814345234348e-05,
+      "loss": 1.4144,
+      "step": 12630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.458368044567466e-05,
+      "loss": 1.4298,
+      "step": 12635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.449924631881277e-05,
+      "loss": 1.3918,
+      "step": 12640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.441484114060865e-05,
+      "loss": 1.4225,
+      "step": 12645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.43304649798894e-05,
+      "loss": 1.4302,
+      "step": 12650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.424611790545862e-05,
+      "loss": 1.4131,
+      "step": 12655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.416179998609604e-05,
+      "loss": 1.4137,
+      "step": 12660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.407751129055772e-05,
+      "loss": 1.384,
+      "step": 12665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.399325188757583e-05,
+      "loss": 1.4397,
+      "step": 12670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.390902184585869e-05,
+      "loss": 1.4054,
+      "step": 12675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.382482123409064e-05,
+      "loss": 1.4166,
+      "step": 12680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.374065012093206e-05,
+      "loss": 1.4157,
+      "step": 12685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.365650857501926e-05,
+      "loss": 1.4305,
+      "step": 12690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5,
+      "learning_rate": 6.357239666496446e-05,
+      "loss": 1.3844,
+      "step": 12695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.348831445935566e-05,
+      "loss": 1.444,
+      "step": 12700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.340426202675669e-05,
+      "loss": 1.45,
+      "step": 12705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.332023943570706e-05,
+      "loss": 1.4239,
+      "step": 12710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.323624675472202e-05,
+      "loss": 1.4083,
+      "step": 12715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.315228405229232e-05,
+      "loss": 1.4053,
+      "step": 12720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.306835139688438e-05,
+      "loss": 1.4305,
+      "step": 12725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.298444885694001e-05,
+      "loss": 1.3939,
+      "step": 12730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.290057650087656e-05,
+      "loss": 1.3813,
+      "step": 12735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.281673439708668e-05,
+      "loss": 1.4162,
+      "step": 12740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.273292261393846e-05,
+      "loss": 1.4266,
+      "step": 12745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.264914121977512e-05,
+      "loss": 1.417,
+      "step": 12750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.256539028291523e-05,
+      "loss": 1.4123,
+      "step": 12755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.248166987165247e-05,
+      "loss": 1.4137,
+      "step": 12760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.239798005425561e-05,
+      "loss": 1.4262,
+      "step": 12765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.231432089896848e-05,
+      "loss": 1.4328,
+      "step": 12770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.223069247400998e-05,
+      "loss": 1.4066,
+      "step": 12775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.214709484757382e-05,
+      "loss": 1.4116,
+      "step": 12780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.206352808782873e-05,
+      "loss": 1.3878,
+      "step": 12785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.197999226291816e-05,
+      "loss": 1.4295,
+      "step": 12790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.189648744096043e-05,
+      "loss": 1.4139,
+      "step": 12795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.181301369004847e-05,
+      "loss": 1.4571,
+      "step": 12800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.172957107824999e-05,
+      "loss": 1.4152,
+      "step": 12805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.164615967360723e-05,
+      "loss": 1.4134,
+      "step": 12810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.156277954413701e-05,
+      "loss": 1.3994,
+      "step": 12815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.147943075783062e-05,
+      "loss": 1.4376,
+      "step": 12820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.139611338265386e-05,
+      "loss": 1.4031,
+      "step": 12825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.131282748654681e-05,
+      "loss": 1.362,
+      "step": 12830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.1229573137424e-05,
+      "loss": 1.4034,
+      "step": 12835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.114635040317414e-05,
+      "loss": 1.3764,
+      "step": 12840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.10631593516602e-05,
+      "loss": 1.4128,
+      "step": 12845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.098000005071933e-05,
+      "loss": 1.4145,
+      "step": 12850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.089687256816276e-05,
+      "loss": 1.4306,
+      "step": 12855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.081377697177576e-05,
+      "loss": 1.4322,
+      "step": 12860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.073071332931768e-05,
+      "loss": 1.4077,
+      "step": 12865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.064768170852169e-05,
+      "loss": 1.3876,
+      "step": 12870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.0564682177094976e-05,
+      "loss": 1.4286,
+      "step": 12875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.048171480271847e-05,
+      "loss": 1.4198,
+      "step": 12880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.0398779653046876e-05,
+      "loss": 1.4129,
+      "step": 12885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.031587679570869e-05,
+      "loss": 1.3972,
+      "step": 12890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.0233006298306024e-05,
+      "loss": 1.3879,
+      "step": 12895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.015016822841465e-05,
+      "loss": 1.4311,
+      "step": 12900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.006736265358381e-05,
+      "loss": 1.4233,
+      "step": 12905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.9984589641336354e-05,
+      "loss": 1.4113,
+      "step": 12910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.9901849259168484e-05,
+      "loss": 1.4208,
+      "step": 12915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.981914157454988e-05,
+      "loss": 1.4341,
+      "step": 12920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.9736466654923476e-05,
+      "loss": 1.3909,
+      "step": 12925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.9653824567705564e-05,
+      "loss": 1.3685,
+      "step": 12930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.9571215380285604e-05,
+      "loss": 1.4367,
+      "step": 12935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.9488639160026274e-05,
+      "loss": 1.4329,
+      "step": 12940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.940609597426332e-05,
+      "loss": 1.4143,
+      "step": 12945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.932358589030562e-05,
+      "loss": 1.407,
+      "step": 12950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.9241108975434976e-05,
+      "loss": 1.4059,
+      "step": 12955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.9158665296906235e-05,
+      "loss": 1.4186,
+      "step": 12960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.9076254921947024e-05,
+      "loss": 1.4117,
+      "step": 12965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.899387791775794e-05,
+      "loss": 1.3814,
+      "step": 12970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.8911534351512276e-05,
+      "loss": 1.4122,
+      "step": 12975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.882922429035611e-05,
+      "loss": 1.4037,
+      "step": 12980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.874694780140817e-05,
+      "loss": 1.435,
+      "step": 12985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.866470495175982e-05,
+      "loss": 1.4324,
+      "step": 12990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.858249580847499e-05,
+      "loss": 1.4104,
+      "step": 12995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.850032043859013e-05,
+      "loss": 1.411,
+      "step": 13000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.841817890911413e-05,
+      "loss": 1.4082,
+      "step": 13005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.8336071287028315e-05,
+      "loss": 1.4199,
+      "step": 13010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.825399763928634e-05,
+      "loss": 1.4118,
+      "step": 13015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.817195803281421e-05,
+      "loss": 1.3896,
+      "step": 13020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.808995253451006e-05,
+      "loss": 1.4292,
+      "step": 13025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.8007981211244276e-05,
+      "loss": 1.4138,
+      "step": 13030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.79260441298594e-05,
+      "loss": 1.3991,
+      "step": 13035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.7844141357170087e-05,
+      "loss": 1.4028,
+      "step": 13040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.776227295996284e-05,
+      "loss": 1.3874,
+      "step": 13045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.768043900499631e-05,
+      "loss": 1.4333,
+      "step": 13050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.759863955900099e-05,
+      "loss": 1.4307,
+      "step": 13055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.751687468867929e-05,
+      "loss": 1.4232,
+      "step": 13060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.74351444607053e-05,
+      "loss": 1.4245,
+      "step": 13065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.7353448941724966e-05,
+      "loss": 1.3847,
+      "step": 13070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.727178819835592e-05,
+      "loss": 1.4309,
+      "step": 13075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.7190162297187475e-05,
+      "loss": 1.4215,
+      "step": 13080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.7108571304780355e-05,
+      "loss": 1.4169,
+      "step": 13085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.702701528766703e-05,
+      "loss": 1.4033,
+      "step": 13090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.694549431235133e-05,
+      "loss": 1.4193,
+      "step": 13095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.6864008445308603e-05,
+      "loss": 1.4367,
+      "step": 13100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.678255775298542e-05,
+      "loss": 1.4123,
+      "step": 13105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.6701142301799784e-05,
+      "loss": 1.4158,
+      "step": 13110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.6619762158140955e-05,
+      "loss": 1.4283,
+      "step": 13115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.6538417388369404e-05,
+      "loss": 1.4124,
+      "step": 13120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.6457108058816674e-05,
+      "loss": 1.4147,
+      "step": 13125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.6375834235785495e-05,
+      "loss": 1.4181,
+      "step": 13130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.62945959855496e-05,
+      "loss": 1.3914,
+      "step": 13135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.6213393374353814e-05,
+      "loss": 1.4017,
+      "step": 13140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.6132226468413715e-05,
+      "loss": 1.3974,
+      "step": 13145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.60510953339159e-05,
+      "loss": 1.4176,
+      "step": 13150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.597000003701779e-05,
+      "loss": 1.4182,
+      "step": 13155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.5888940643847574e-05,
+      "loss": 1.4126,
+      "step": 13160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.580791722050408e-05,
+      "loss": 1.41,
+      "step": 13165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.5726929833056954e-05,
+      "loss": 1.3945,
+      "step": 13170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.5645978547546284e-05,
+      "loss": 1.3876,
+      "step": 13175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.5565063429982865e-05,
+      "loss": 1.4207,
+      "step": 13180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.5484184546347983e-05,
+      "loss": 1.4061,
+      "step": 13185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.540334196259326e-05,
+      "loss": 1.3876,
+      "step": 13190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.532253574464083e-05,
+      "loss": 1.4197,
+      "step": 13195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.5241765958383154e-05,
+      "loss": 1.4209,
+      "step": 13200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.516103266968299e-05,
+      "loss": 1.3889,
+      "step": 13205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.508033594437325e-05,
+      "loss": 1.4103,
+      "step": 13210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.4999675848257147e-05,
+      "loss": 1.3954,
+      "step": 13215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.491905244710796e-05,
+      "loss": 1.4404,
+      "step": 13220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.48384658066691e-05,
+      "loss": 1.3895,
+      "step": 13225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.47579159926539e-05,
+      "loss": 1.4484,
+      "step": 13230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.467740307074574e-05,
+      "loss": 1.4115,
+      "step": 13235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.459692710659792e-05,
+      "loss": 1.4206,
+      "step": 13240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.451648816583362e-05,
+      "loss": 1.429,
+      "step": 13245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.443608631404573e-05,
+      "loss": 1.4297,
+      "step": 13250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.435572161679698e-05,
+      "loss": 1.3839,
+      "step": 13255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.42753941396198e-05,
+      "loss": 1.3952,
+      "step": 13260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.419510394801628e-05,
+      "loss": 1.3986,
+      "step": 13265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.411485110745802e-05,
+      "loss": 1.4064,
+      "step": 13270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.4034635683386245e-05,
+      "loss": 1.4064,
+      "step": 13275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.395445774121166e-05,
+      "loss": 1.3951,
+      "step": 13280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.387431734631443e-05,
+      "loss": 1.4237,
+      "step": 13285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.379421456404397e-05,
+      "loss": 1.4112,
+      "step": 13290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.371414945971918e-05,
+      "loss": 1.4125,
+      "step": 13295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.3634122098628146e-05,
+      "loss": 1.4689,
+      "step": 13300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.3554132546028294e-05,
+      "loss": 1.3824,
+      "step": 13305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.3474180867146004e-05,
+      "loss": 1.4219,
+      "step": 13310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.339426712717697e-05,
+      "loss": 1.4107,
+      "step": 13315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.331439139128587e-05,
+      "loss": 1.4263,
+      "step": 13320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.323455372460644e-05,
+      "loss": 1.3911,
+      "step": 13325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.315475419224124e-05,
+      "loss": 1.4176,
+      "step": 13330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.3074992859261895e-05,
+      "loss": 1.4085,
+      "step": 13335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.299526979070879e-05,
+      "loss": 1.3893,
+      "step": 13340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.2915585051591196e-05,
+      "loss": 1.3686,
+      "step": 13345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.2835938706886966e-05,
+      "loss": 1.3746,
+      "step": 13350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.275633082154279e-05,
+      "loss": 1.3683,
+      "step": 13355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.2676761460473934e-05,
+      "loss": 1.4154,
+      "step": 13360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.259723068856434e-05,
+      "loss": 1.387,
+      "step": 13365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.251773857066629e-05,
+      "loss": 1.3847,
+      "step": 13370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.243828517160072e-05,
+      "loss": 1.3706,
+      "step": 13375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.235887055615696e-05,
+      "loss": 1.4202,
+      "step": 13380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.227949478909265e-05,
+      "loss": 1.3961,
+      "step": 13385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.2200157935133865e-05,
+      "loss": 1.4123,
+      "step": 13390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.2120860058974786e-05,
+      "loss": 1.4374,
+      "step": 13395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.204160122527795e-05,
+      "loss": 1.3853,
+      "step": 13400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.196238149867398e-05,
+      "loss": 1.4036,
+      "step": 13405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.188320094376172e-05,
+      "loss": 1.4003,
+      "step": 13410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.180405962510789e-05,
+      "loss": 1.4341,
+      "step": 13415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.172495760724736e-05,
+      "loss": 1.4119,
+      "step": 13420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.1645894954682896e-05,
+      "loss": 1.3961,
+      "step": 13425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.156687173188521e-05,
+      "loss": 1.3973,
+      "step": 13430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.148788800329278e-05,
+      "loss": 1.411,
+      "step": 13435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.140894383331196e-05,
+      "loss": 1.4306,
+      "step": 13440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.133003928631679e-05,
+      "loss": 1.3455,
+      "step": 13445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.1251174426649076e-05,
+      "loss": 1.4141,
+      "step": 13450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.117234931861813e-05,
+      "loss": 1.3502,
+      "step": 13455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.109356402650096e-05,
+      "loss": 1.4024,
+      "step": 13460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.1014818614542116e-05,
+      "loss": 1.415,
+      "step": 13465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.0936113146953525e-05,
+      "loss": 1.4265,
+      "step": 13470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.085744768791465e-05,
+      "loss": 1.3924,
+      "step": 13475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.0778822301572226e-05,
+      "loss": 1.3742,
+      "step": 13480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.070023705204041e-05,
+      "loss": 1.4227,
+      "step": 13485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.062169200340058e-05,
+      "loss": 1.4204,
+      "step": 13490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.054318721970137e-05,
+      "loss": 1.3805,
+      "step": 13495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.046472276495848e-05,
+      "loss": 1.3879,
+      "step": 13500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.038629870315486e-05,
+      "loss": 1.3919,
+      "step": 13505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.030791509824041e-05,
+      "loss": 1.4157,
+      "step": 13510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.0229572014132156e-05,
+      "loss": 1.4404,
+      "step": 13515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.0151269514713927e-05,
+      "loss": 1.4115,
+      "step": 13520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.007300766383659e-05,
+      "loss": 1.4164,
+      "step": 13525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.999478652531782e-05,
+      "loss": 1.3713,
+      "step": 13530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.99166061629421e-05,
+      "loss": 1.3609,
+      "step": 13535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.9838466640460627e-05,
+      "loss": 1.3923,
+      "step": 13540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.976036802159133e-05,
+      "loss": 1.4069,
+      "step": 13545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.968231037001879e-05,
+      "loss": 1.3967,
+      "step": 13550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.96042937493942e-05,
+      "loss": 1.3987,
+      "step": 13555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5,
+      "learning_rate": 4.95263182233352e-05,
+      "loss": 1.4095,
+      "step": 13560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.9448383855426006e-05,
+      "loss": 1.3958,
+      "step": 13565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.937049070921727e-05,
+      "loss": 1.4014,
+      "step": 13570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.9292638848226024e-05,
+      "loss": 1.4096,
+      "step": 13575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.9214828335935556e-05,
+      "loss": 1.4241,
+      "step": 13580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.913705923579556e-05,
+      "loss": 1.4062,
+      "step": 13585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.905933161122187e-05,
+      "loss": 1.3818,
+      "step": 13590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.89816455255966e-05,
+      "loss": 1.4079,
+      "step": 13595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.890400104226782e-05,
+      "loss": 1.4027,
+      "step": 13600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.882639822454983e-05,
+      "loss": 1.3885,
+      "step": 13605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.87488371357229e-05,
+      "loss": 1.4119,
+      "step": 13610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.867131783903333e-05,
+      "loss": 1.4109,
+      "step": 13615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.859384039769319e-05,
+      "loss": 1.4487,
+      "step": 13620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.851640487488057e-05,
+      "loss": 1.404,
+      "step": 13625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.8439011333739314e-05,
+      "loss": 1.4212,
+      "step": 13630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.836165983737909e-05,
+      "loss": 1.3863,
+      "step": 13635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.828435044887516e-05,
+      "loss": 1.3797,
+      "step": 13640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.820708323126856e-05,
+      "loss": 1.4255,
+      "step": 13645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.81298582475659e-05,
+      "loss": 1.3684,
+      "step": 13650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.805267556073938e-05,
+      "loss": 1.3921,
+      "step": 13655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.797553523372663e-05,
+      "loss": 1.419,
+      "step": 13660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.7898437329430815e-05,
+      "loss": 1.4267,
+      "step": 13665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.7821381910720484e-05,
+      "loss": 1.3806,
+      "step": 13670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.774436904042959e-05,
+      "loss": 1.4022,
+      "step": 13675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.766739878135725e-05,
+      "loss": 1.4005,
+      "step": 13680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.759047119626798e-05,
+      "loss": 1.4232,
+      "step": 13685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.751358634789143e-05,
+      "loss": 1.4182,
+      "step": 13690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.743674429892245e-05,
+      "loss": 1.4279,
+      "step": 13695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.73599451120209e-05,
+      "loss": 1.3974,
+      "step": 13700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.728318884981175e-05,
+      "loss": 1.3948,
+      "step": 13705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.7206475574884976e-05,
+      "loss": 1.4153,
+      "step": 13710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.712980534979553e-05,
+      "loss": 1.394,
+      "step": 13715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.7053178237063135e-05,
+      "loss": 1.3956,
+      "step": 13720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.697659429917246e-05,
+      "loss": 1.4027,
+      "step": 13725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.690005359857297e-05,
+      "loss": 1.3886,
+      "step": 13730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6823556197678865e-05,
+      "loss": 1.4122,
+      "step": 13735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.674710215886895e-05,
+      "loss": 1.408,
+      "step": 13740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.667069154448679e-05,
+      "loss": 1.4116,
+      "step": 13745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.659432441684047e-05,
+      "loss": 1.3956,
+      "step": 13750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.6518000838202694e-05,
+      "loss": 1.4042,
+      "step": 13755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.6441720870810545e-05,
+      "loss": 1.4179,
+      "step": 13760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.636548457686557e-05,
+      "loss": 1.3851,
+      "step": 13765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.628929201853375e-05,
+      "loss": 1.4158,
+      "step": 13770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.621314325794539e-05,
+      "loss": 1.3648,
+      "step": 13775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.613703835719511e-05,
+      "loss": 1.3847,
+      "step": 13780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.606097737834163e-05,
+      "loss": 1.4183,
+      "step": 13785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.5984960383408005e-05,
+      "loss": 1.3988,
+      "step": 13790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.590898743438138e-05,
+      "loss": 1.4255,
+      "step": 13795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.5833058593212984e-05,
+      "loss": 1.4159,
+      "step": 13800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.575717392181801e-05,
+      "loss": 1.4114,
+      "step": 13805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.568133348207572e-05,
+      "loss": 1.4001,
+      "step": 13810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.5605537335829275e-05,
+      "loss": 1.3891,
+      "step": 13815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.5529785544885715e-05,
+      "loss": 1.4303,
+      "step": 13820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.545407817101598e-05,
+      "loss": 1.4256,
+      "step": 13825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.5378415275954634e-05,
+      "loss": 1.4324,
+      "step": 13830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.53027969214001e-05,
+      "loss": 1.4147,
+      "step": 13835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.5227223169014456e-05,
+      "loss": 1.4096,
+      "step": 13840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.5151694080423414e-05,
+      "loss": 1.417,
+      "step": 13845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.50762097172162e-05,
+      "loss": 1.4225,
+      "step": 13850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.500077014094566e-05,
+      "loss": 1.4455,
+      "step": 13855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.492537541312805e-05,
+      "loss": 1.415,
+      "step": 13860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.485002559524314e-05,
+      "loss": 1.4214,
+      "step": 13865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.477472074873396e-05,
+      "loss": 1.424,
+      "step": 13870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.469946093500694e-05,
+      "loss": 1.3948,
+      "step": 13875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.4624246215431796e-05,
+      "loss": 1.3954,
+      "step": 13880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.4549076651341493e-05,
+      "loss": 1.449,
+      "step": 13885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.4473952304032065e-05,
+      "loss": 1.3918,
+      "step": 13890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.439887323476277e-05,
+      "loss": 1.4122,
+      "step": 13895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.432383950475595e-05,
+      "loss": 1.4061,
+      "step": 13900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.4248851175196956e-05,
+      "loss": 1.4182,
+      "step": 13905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.4173908307234045e-05,
+      "loss": 1.4331,
+      "step": 13910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.40990109619785e-05,
+      "loss": 1.4382,
+      "step": 13915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.402415920050447e-05,
+      "loss": 1.406,
+      "step": 13920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.394935308384893e-05,
+      "loss": 1.4181,
+      "step": 13925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.387459267301155e-05,
+      "loss": 1.4027,
+      "step": 13930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.379987802895483e-05,
+      "loss": 1.3861,
+      "step": 13935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.3725209212603925e-05,
+      "loss": 1.402,
+      "step": 13940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.3650586284846636e-05,
+      "loss": 1.4507,
+      "step": 13945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.357600930653327e-05,
+      "loss": 1.4307,
+      "step": 13950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.350147833847674e-05,
+      "loss": 1.4185,
+      "step": 13955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.3426993441452414e-05,
+      "loss": 1.3886,
+      "step": 13960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.335255467619814e-05,
+      "loss": 1.402,
+      "step": 13965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.3278162103414033e-05,
+      "loss": 1.4353,
+      "step": 13970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.320381578376264e-05,
+      "loss": 1.4512,
+      "step": 13975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.312951577786876e-05,
+      "loss": 1.4378,
+      "step": 13980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.305526214631948e-05,
+      "loss": 1.4088,
+      "step": 13985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.2981054949663926e-05,
+      "loss": 1.3837,
+      "step": 13990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.290689424841351e-05,
+      "loss": 1.3985,
+      "step": 13995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.283278010304167e-05,
+      "loss": 1.411,
+      "step": 14000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.2758712573983915e-05,
+      "loss": 1.3721,
+      "step": 14005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.268469172163764e-05,
+      "loss": 1.4247,
+      "step": 14010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.261071760636228e-05,
+      "loss": 1.4248,
+      "step": 14015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.2536790288479135e-05,
+      "loss": 1.4128,
+      "step": 14020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.246290982827137e-05,
+      "loss": 1.385,
+      "step": 14025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.238907628598384e-05,
+      "loss": 1.3891,
+      "step": 14030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.231528972182324e-05,
+      "loss": 1.4083,
+      "step": 14035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.2241550195957924e-05,
+      "loss": 1.4192,
+      "step": 14040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.2167857768517935e-05,
+      "loss": 1.4116,
+      "step": 14045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.2094212499594785e-05,
+      "loss": 1.3968,
+      "step": 14050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.2020614449241705e-05,
+      "loss": 1.4164,
+      "step": 14055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.194706367747323e-05,
+      "loss": 1.3963,
+      "step": 14060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.187356024426549e-05,
+      "loss": 1.4072,
+      "step": 14065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.1800104209556e-05,
+      "loss": 1.4131,
+      "step": 14070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.1726695633243527e-05,
+      "loss": 1.4076,
+      "step": 14075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.165333457518823e-05,
+      "loss": 1.3496,
+      "step": 14080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.1580021095211486e-05,
+      "loss": 1.4122,
+      "step": 14085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.150675525309593e-05,
+      "loss": 1.4152,
+      "step": 14090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.1433537108585216e-05,
+      "loss": 1.4124,
+      "step": 14095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.1360366721384234e-05,
+      "loss": 1.4217,
+      "step": 14100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.128724415115889e-05,
+      "loss": 1.4085,
+      "step": 14105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.121416945753611e-05,
+      "loss": 1.3978,
+      "step": 14110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.114114270010372e-05,
+      "loss": 1.3935,
+      "step": 14115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.106816393841052e-05,
+      "loss": 1.4313,
+      "step": 14120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.099523323196616e-05,
+      "loss": 1.4194,
+      "step": 14125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.092235064024111e-05,
+      "loss": 1.3693,
+      "step": 14130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.0849516222666564e-05,
+      "loss": 1.4015,
+      "step": 14135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.077673003863446e-05,
+      "loss": 1.4321,
+      "step": 14140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.0703992147497425e-05,
+      "loss": 1.397,
+      "step": 14145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.063130260856872e-05,
+      "loss": 1.4254,
+      "step": 14150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.055866148112208e-05,
+      "loss": 1.3843,
+      "step": 14155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.0486068824391856e-05,
+      "loss": 1.3909,
+      "step": 14160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.041352469757283e-05,
+      "loss": 1.4008,
+      "step": 14165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.034102915982031e-05,
+      "loss": 1.4209,
+      "step": 14170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.026858227024978e-05,
+      "loss": 1.4422,
+      "step": 14175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.0196184087937235e-05,
+      "loss": 1.424,
+      "step": 14180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.012383467191889e-05,
+      "loss": 1.4286,
+      "step": 14185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.005153408119123e-05,
+      "loss": 1.4388,
+      "step": 14190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.9979282374710824e-05,
+      "loss": 1.3725,
+      "step": 14195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.9907079611394485e-05,
+      "loss": 1.4091,
+      "step": 14200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.983492585011906e-05,
+      "loss": 1.3604,
+      "step": 14205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.9762821149721485e-05,
+      "loss": 1.4136,
+      "step": 14210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.9690765568998665e-05,
+      "loss": 1.3978,
+      "step": 14215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.9618759166707396e-05,
+      "loss": 1.375,
+      "step": 14220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.9546802001564454e-05,
+      "loss": 1.4229,
+      "step": 14225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.9474894132246435e-05,
+      "loss": 1.4022,
+      "step": 14230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.940303561738977e-05,
+      "loss": 1.4145,
+      "step": 14235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.933122651559054e-05,
+      "loss": 1.4006,
+      "step": 14240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.925946688540464e-05,
+      "loss": 1.4169,
+      "step": 14245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.918775678534759e-05,
+      "loss": 1.4083,
+      "step": 14250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.911609627389453e-05,
+      "loss": 1.4037,
+      "step": 14255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.904448540948012e-05,
+      "loss": 1.4371,
+      "step": 14260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.897292425049859e-05,
+      "loss": 1.3692,
+      "step": 14265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.89014128553036e-05,
+      "loss": 1.4395,
+      "step": 14270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.8829951282208297e-05,
+      "loss": 1.3866,
+      "step": 14275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.875853958948508e-05,
+      "loss": 1.4246,
+      "step": 14280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.868717783536578e-05,
+      "loss": 1.4123,
+      "step": 14285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.861586607804147e-05,
+      "loss": 1.4189,
+      "step": 14290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.8544604375662495e-05,
+      "loss": 1.4224,
+      "step": 14295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.847339278633827e-05,
+      "loss": 1.4446,
+      "step": 14300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.8402231368137454e-05,
+      "loss": 1.3983,
+      "step": 14305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.8331120179087754e-05,
+      "loss": 1.4149,
+      "step": 14310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.8260059277175965e-05,
+      "loss": 1.3856,
+      "step": 14315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.818904872034777e-05,
+      "loss": 1.4143,
+      "step": 14320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.8118088566507884e-05,
+      "loss": 1.4227,
+      "step": 14325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.804717887351991e-05,
+      "loss": 1.412,
+      "step": 14330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.797631969920633e-05,
+      "loss": 1.424,
+      "step": 14335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.7905511101348334e-05,
+      "loss": 1.432,
+      "step": 14340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.7834753137685955e-05,
+      "loss": 1.4125,
+      "step": 14345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.776404586591794e-05,
+      "loss": 1.4444,
+      "step": 14350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.769338934370163e-05,
+      "loss": 1.4182,
+      "step": 14355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.762278362865308e-05,
+      "loss": 1.4039,
+      "step": 14360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.755222877834679e-05,
+      "loss": 1.4062,
+      "step": 14365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.7481724850315894e-05,
+      "loss": 1.3981,
+      "step": 14370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.741127190205196e-05,
+      "loss": 1.3981,
+      "step": 14375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.734086999100502e-05,
+      "loss": 1.4209,
+      "step": 14380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.7270519174583404e-05,
+      "loss": 1.3797,
+      "step": 14385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.7200219510153845e-05,
+      "loss": 1.406,
+      "step": 14390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.7129971055041345e-05,
+      "loss": 1.4059,
+      "step": 14395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.705977386652921e-05,
+      "loss": 1.3911,
+      "step": 14400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.69896280018588e-05,
+      "loss": 1.438,
+      "step": 14405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.6919533518229734e-05,
+      "loss": 1.3953,
+      "step": 14410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.6849490472799716e-05,
+      "loss": 1.4092,
+      "step": 14415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.677949892268453e-05,
+      "loss": 1.4013,
+      "step": 14420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.670955892495787e-05,
+      "loss": 1.4047,
+      "step": 14425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.663967053665147e-05,
+      "loss": 1.3838,
+      "step": 14430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.6569833814754995e-05,
+      "loss": 1.3798,
+      "step": 14435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.650004881621596e-05,
+      "loss": 1.4116,
+      "step": 14440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.6430315597939636e-05,
+      "loss": 1.3809,
+      "step": 14445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.636063421678917e-05,
+      "loss": 1.4039,
+      "step": 14450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.629100472958538e-05,
+      "loss": 1.4236,
+      "step": 14455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.6221427193106814e-05,
+      "loss": 1.3737,
+      "step": 14460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.615190166408959e-05,
+      "loss": 1.391,
+      "step": 14465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.608242819922746e-05,
+      "loss": 1.4241,
+      "step": 14470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.6013006855171726e-05,
+      "loss": 1.4078,
+      "step": 14475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.5943637688531216e-05,
+      "loss": 1.4076,
+      "step": 14480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.58743207558721e-05,
+      "loss": 1.3783,
+      "step": 14485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.580505611371806e-05,
+      "loss": 1.4014,
+      "step": 14490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.573584381855012e-05,
+      "loss": 1.3963,
+      "step": 14495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.566668392680662e-05,
+      "loss": 1.4465,
+      "step": 14500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.5597576494883086e-05,
+      "loss": 1.3737,
+      "step": 14505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.552852157913238e-05,
+      "loss": 1.4044,
+      "step": 14510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.545951923586448e-05,
+      "loss": 1.4472,
+      "step": 14515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.539056952134655e-05,
+      "loss": 1.3723,
+      "step": 14520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.532167249180271e-05,
+      "loss": 1.4226,
+      "step": 14525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.525282820341428e-05,
+      "loss": 1.3928,
+      "step": 14530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.5184036712319444e-05,
+      "loss": 1.4194,
+      "step": 14535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.5115298074613466e-05,
+      "loss": 1.4332,
+      "step": 14540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.504661234634834e-05,
+      "loss": 1.3965,
+      "step": 14545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.497797958353305e-05,
+      "loss": 1.4264,
+      "step": 14550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.490939984213334e-05,
+      "loss": 1.3683,
+      "step": 14555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.484087317807176e-05,
+      "loss": 1.4378,
+      "step": 14560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.477239964722748e-05,
+      "loss": 1.3627,
+      "step": 14565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.470397930543645e-05,
+      "loss": 1.3664,
+      "step": 14570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.4635612208491194e-05,
+      "loss": 1.3869,
+      "step": 14575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.456729841214083e-05,
+      "loss": 1.3984,
+      "step": 14580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.4499037972091064e-05,
+      "loss": 1.3961,
+      "step": 14585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.443083094400395e-05,
+      "loss": 1.3964,
+      "step": 14590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.4362677383498123e-05,
+      "loss": 1.4334,
+      "step": 14595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.429457734614857e-05,
+      "loss": 1.3997,
+      "step": 14600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.422653088748668e-05,
+      "loss": 1.382,
+      "step": 14605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.4158538063000046e-05,
+      "loss": 1.4022,
+      "step": 14610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.409059892813261e-05,
+      "loss": 1.3996,
+      "step": 14615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.402271353828452e-05,
+      "loss": 1.444,
+      "step": 14620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.3954881948812125e-05,
+      "loss": 1.4276,
+      "step": 14625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.38871042150278e-05,
+      "loss": 1.3799,
+      "step": 14630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.381938039220011e-05,
+      "loss": 1.4047,
+      "step": 14635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.3751710535553615e-05,
+      "loss": 1.3985,
+      "step": 14640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.368409470026892e-05,
+      "loss": 1.3995,
+      "step": 14645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.3616532941482494e-05,
+      "loss": 1.4014,
+      "step": 14650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.354902531428673e-05,
+      "loss": 1.4169,
+      "step": 14655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.3481571873729924e-05,
+      "loss": 1.4251,
+      "step": 14660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.341417267481616e-05,
+      "loss": 1.4041,
+      "step": 14665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.334682777250534e-05,
+      "loss": 1.4032,
+      "step": 14670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.3279537221712975e-05,
+      "loss": 1.4211,
+      "step": 14675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.321230107731035e-05,
+      "loss": 1.3886,
+      "step": 14680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.314511939412438e-05,
+      "loss": 1.3968,
+      "step": 14685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.307799222693756e-05,
+      "loss": 1.4073,
+      "step": 14690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.301091963048788e-05,
+      "loss": 1.3847,
+      "step": 14695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.294390165946889e-05,
+      "loss": 1.4143,
+      "step": 14700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.287693836852959e-05,
+      "loss": 1.3882,
+      "step": 14705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.281002981227439e-05,
+      "loss": 1.3611,
+      "step": 14710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.2743176045263024e-05,
+      "loss": 1.3952,
+      "step": 14715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.2676377122010605e-05,
+      "loss": 1.4149,
+      "step": 14720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.260963309698749e-05,
+      "loss": 1.3637,
+      "step": 14725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.254294402461933e-05,
+      "loss": 1.4297,
+      "step": 14730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.2476309959286846e-05,
+      "loss": 1.3733,
+      "step": 14735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.240973095532601e-05,
+      "loss": 1.4047,
+      "step": 14740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.2343207067027856e-05,
+      "loss": 1.4128,
+      "step": 14745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.227673834863852e-05,
+      "loss": 1.4404,
+      "step": 14750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.221032485435904e-05,
+      "loss": 1.4148,
+      "step": 14755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.214396663834553e-05,
+      "loss": 1.4185,
+      "step": 14760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.2077663754708983e-05,
+      "loss": 1.4212,
+      "step": 14765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.201141625751532e-05,
+      "loss": 1.4082,
+      "step": 14770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.194522420078518e-05,
+      "loss": 1.3706,
+      "step": 14775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.187908763849412e-05,
+      "loss": 1.4007,
+      "step": 14780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.181300662457237e-05,
+      "loss": 1.4022,
+      "step": 14785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.1746981212904944e-05,
+      "loss": 1.4264,
+      "step": 14790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.168101145733139e-05,
+      "loss": 1.4189,
+      "step": 14795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.161509741164596e-05,
+      "loss": 1.4334,
+      "step": 14800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.1549239129597484e-05,
+      "loss": 1.4155,
+      "step": 14805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.148343666488931e-05,
+      "loss": 1.3467,
+      "step": 14810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.141769007117921e-05,
+      "loss": 1.4213,
+      "step": 14815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.135199940207947e-05,
+      "loss": 1.3895,
+      "step": 14820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.1286364711156734e-05,
+      "loss": 1.4077,
+      "step": 14825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.1220786051932064e-05,
+      "loss": 1.4263,
+      "step": 14830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.1155263477880703e-05,
+      "loss": 1.3859,
+      "step": 14835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.108979704243228e-05,
+      "loss": 1.4295,
+      "step": 14840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.1024386798970586e-05,
+      "loss": 1.3763,
+      "step": 14845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.0959032800833657e-05,
+      "loss": 1.4086,
+      "step": 14850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.089373510131354e-05,
+      "loss": 1.4125,
+      "step": 14855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.0828493753656495e-05,
+      "loss": 1.4476,
+      "step": 14860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.076330881106278e-05,
+      "loss": 1.4296,
+      "step": 14865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.069818032668668e-05,
+      "loss": 1.434,
+      "step": 14870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.0633108353636376e-05,
+      "loss": 1.4098,
+      "step": 14875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.056809294497406e-05,
+      "loss": 1.3864,
+      "step": 14880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.050313415371573e-05,
+      "loss": 1.3848,
+      "step": 14885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.0438232032831292e-05,
+      "loss": 1.3968,
+      "step": 14890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.0373386635244327e-05,
+      "loss": 1.4014,
+      "step": 14895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.0308598013832256e-05,
+      "loss": 1.3744,
+      "step": 14900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.0243866221426166e-05,
+      "loss": 1.3868,
+      "step": 14905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.0179191310810838e-05,
+      "loss": 1.3965,
+      "step": 14910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.0114573334724592e-05,
+      "loss": 1.4171,
+      "step": 14915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.005001234585939e-05,
+      "loss": 1.4186,
+      "step": 14920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9985508396860717e-05,
+      "loss": 1.4451,
+      "step": 14925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.9921061540327545e-05,
+      "loss": 1.4018,
+      "step": 14930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.9856671828812244e-05,
+      "loss": 1.3988,
+      "step": 14935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.9792339314820662e-05,
+      "loss": 1.3915,
+      "step": 14940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.972806405081191e-05,
+      "loss": 1.4495,
+      "step": 14945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.96638460891985e-05,
+      "loss": 1.4278,
+      "step": 14950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.9599685482346218e-05,
+      "loss": 1.3901,
+      "step": 14955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.9535582282573982e-05,
+      "loss": 1.3876,
+      "step": 14960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.947153654215402e-05,
+      "loss": 1.4109,
+      "step": 14965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.940754831331163e-05,
+      "loss": 1.3847,
+      "step": 14970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.9343617648225273e-05,
+      "loss": 1.3946,
+      "step": 14975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.927974459902637e-05,
+      "loss": 1.4048,
+      "step": 14980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.9215929217799454e-05,
+      "loss": 1.4128,
+      "step": 14985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.9152171556581998e-05,
+      "loss": 1.3907,
+      "step": 14990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.9088471667364447e-05,
+      "loss": 1.4094,
+      "step": 14995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9024829602090033e-05,
+      "loss": 1.4178,
+      "step": 15000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.8961245412654936e-05,
+      "loss": 1.4357,
+      "step": 15005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.889771915090812e-05,
+      "loss": 1.433,
+      "step": 15010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.883425086865129e-05,
+      "loss": 1.4404,
+      "step": 15015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.8770840617638927e-05,
+      "loss": 1.3897,
+      "step": 15020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.8707488449578068e-05,
+      "loss": 1.4184,
+      "step": 15025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.8644194416128523e-05,
+      "loss": 1.4287,
+      "step": 15030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.8580958568902616e-05,
+      "loss": 1.409,
+      "step": 15035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.85177809594653e-05,
+      "loss": 1.4083,
+      "step": 15040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.8454661639333923e-05,
+      "loss": 1.3748,
+      "step": 15045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.839160065997839e-05,
+      "loss": 1.4489,
+      "step": 15050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.832859807282102e-05,
+      "loss": 1.3881,
+      "step": 15055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.8265653929236537e-05,
+      "loss": 1.4145,
+      "step": 15060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.8202768280551894e-05,
+      "loss": 1.4271,
+      "step": 15065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.813994117804648e-05,
+      "loss": 1.3984,
+      "step": 15070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.807717267295189e-05,
+      "loss": 1.3794,
+      "step": 15075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8014462816451958e-05,
+      "loss": 1.4091,
+      "step": 15080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7951811659682625e-05,
+      "loss": 1.4287,
+      "step": 15085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7889219253732046e-05,
+      "loss": 1.4234,
+      "step": 15090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7826685649640428e-05,
+      "loss": 1.4419,
+      "step": 15095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.7764210898400066e-05,
+      "loss": 1.4219,
+      "step": 15100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.770179505095518e-05,
+      "loss": 1.4274,
+      "step": 15105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.7639438158202037e-05,
+      "loss": 1.4231,
+      "step": 15110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.757714027098882e-05,
+      "loss": 1.3985,
+      "step": 15115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.7514901440115615e-05,
+      "loss": 1.3931,
+      "step": 15120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.745272171633424e-05,
+      "loss": 1.4121,
+      "step": 15125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.7390601150348437e-05,
+      "loss": 1.3921,
+      "step": 15130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7328539792813668e-05,
+      "loss": 1.382,
+      "step": 15135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.7266537694337147e-05,
+      "loss": 1.42,
+      "step": 15140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7204594905477655e-05,
+      "loss": 1.3715,
+      "step": 15145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.714271147674572e-05,
+      "loss": 1.3723,
+      "step": 15150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.7080887458603432e-05,
+      "loss": 1.4189,
+      "step": 15155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7019122901464477e-05,
+      "loss": 1.4318,
+      "step": 15160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.6957417855693934e-05,
+      "loss": 1.3811,
+      "step": 15165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.6895772371608473e-05,
+      "loss": 1.4058,
+      "step": 15170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6834186499476145e-05,
+      "loss": 1.3842,
+      "step": 15175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.677266028951645e-05,
+      "loss": 1.3844,
+      "step": 15180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.67111937919001e-05,
+      "loss": 1.4222,
+      "step": 15185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6649787056749254e-05,
+      "loss": 1.4068,
+      "step": 15190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.658844013413727e-05,
+      "loss": 1.4152,
+      "step": 15195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.6527153074088797e-05,
+      "loss": 1.3608,
+      "step": 15200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.6465925926579548e-05,
+      "loss": 1.4029,
+      "step": 15205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.6404758741536505e-05,
+      "loss": 1.386,
+      "step": 15210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.634365156883768e-05,
+      "loss": 1.4222,
+      "step": 15215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.628260445831222e-05,
+      "loss": 1.3958,
+      "step": 15220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.622161745974019e-05,
+      "loss": 1.4211,
+      "step": 15225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.6160690622852746e-05,
+      "loss": 1.4084,
+      "step": 15230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.6099823997331886e-05,
+      "loss": 1.3864,
+      "step": 15235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6039017632810582e-05,
+      "loss": 1.4074,
+      "step": 15240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.597827157887267e-05,
+      "loss": 1.4254,
+      "step": 15245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.5917585885052742e-05,
+      "loss": 1.3996,
+      "step": 15250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.585696060083621e-05,
+      "loss": 1.4215,
+      "step": 15255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.5796395775659243e-05,
+      "loss": 1.3848,
+      "step": 15260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.5735891458908713e-05,
+      "loss": 1.3692,
+      "step": 15265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.5675447699922084e-05,
+      "loss": 1.4185,
+      "step": 15270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.5615064547987487e-05,
+      "loss": 1.4189,
+      "step": 15275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.555474205234366e-05,
+      "loss": 1.353,
+      "step": 15280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.5494480262179855e-05,
+      "loss": 1.3896,
+      "step": 15285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.543427922663576e-05,
+      "loss": 1.4108,
+      "step": 15290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.537413899480161e-05,
+      "loss": 1.3963,
+      "step": 15295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.5314059615718034e-05,
+      "loss": 1.4301,
+      "step": 15300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.525404113837605e-05,
+      "loss": 1.3944,
+      "step": 15305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.5194083611716935e-05,
+      "loss": 1.4192,
+      "step": 15310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5134187084632356e-05,
+      "loss": 1.4197,
+      "step": 15315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.507435160596422e-05,
+      "loss": 1.4077,
+      "step": 15320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.5014577224504642e-05,
+      "loss": 1.3889,
+      "step": 15325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4954863988995892e-05,
+      "loss": 1.4001,
+      "step": 15330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.4895211948130394e-05,
+      "loss": 1.4222,
+      "step": 15335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.48356211505507e-05,
+      "loss": 1.3745,
+      "step": 15340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.4776091644849432e-05,
+      "loss": 1.4021,
+      "step": 15345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.4716623479569136e-05,
+      "loss": 1.4166,
+      "step": 15350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4657216703202435e-05,
+      "loss": 1.4277,
+      "step": 15355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.459787136419186e-05,
+      "loss": 1.3755,
+      "step": 15360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.4538587510929878e-05,
+      "loss": 1.4201,
+      "step": 15365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4479365191758717e-05,
+      "loss": 1.4043,
+      "step": 15370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4420204454970542e-05,
+      "loss": 1.4268,
+      "step": 15375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.4361105348807256e-05,
+      "loss": 1.4083,
+      "step": 15380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.430206792146049e-05,
+      "loss": 1.3965,
+      "step": 15385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.424309222107164e-05,
+      "loss": 1.4135,
+      "step": 15390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.418417829573165e-05,
+      "loss": 1.431,
+      "step": 15395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.4125326193481213e-05,
+      "loss": 1.3714,
+      "step": 15400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4066535962310553e-05,
+      "loss": 1.4141,
+      "step": 15405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.4007807650159464e-05,
+      "loss": 1.425,
+      "step": 15410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.394914130491719e-05,
+      "loss": 1.4037,
+      "step": 15415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3890536974422518e-05,
+      "loss": 1.381,
+      "step": 15420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3831994706463623e-05,
+      "loss": 1.401,
+      "step": 15425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3773514548778132e-05,
+      "loss": 1.4207,
+      "step": 15430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3715096549052908e-05,
+      "loss": 1.4072,
+      "step": 15435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.3656740754924233e-05,
+      "loss": 1.3835,
+      "step": 15440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.3598447213977625e-05,
+      "loss": 1.3809,
+      "step": 15445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.354021597374787e-05,
+      "loss": 1.3715,
+      "step": 15450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3482047081718884e-05,
+      "loss": 1.3965,
+      "step": 15455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.342394058532378e-05,
+      "loss": 1.3943,
+      "step": 15460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.336589653194482e-05,
+      "loss": 1.4223,
+      "step": 15465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.3307914968913347e-05,
+      "loss": 1.3886,
+      "step": 15470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.324999594350965e-05,
+      "loss": 1.3564,
+      "step": 15475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.319213950296314e-05,
+      "loss": 1.4224,
+      "step": 15480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.313434569445213e-05,
+      "loss": 1.392,
+      "step": 15485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.3076614565103916e-05,
+      "loss": 1.4152,
+      "step": 15490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3018946161994594e-05,
+      "loss": 1.4077,
+      "step": 15495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2961340532149177e-05,
+      "loss": 1.4043,
+      "step": 15500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.2903797722541487e-05,
+      "loss": 1.3987,
+      "step": 15505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.2846317780094127e-05,
+      "loss": 1.3903,
+      "step": 15510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.2788900751678367e-05,
+      "loss": 1.4161,
+      "step": 15515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2731546684114247e-05,
+      "loss": 1.3782,
+      "step": 15520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2674255624170472e-05,
+      "loss": 1.3911,
+      "step": 15525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.261702761856429e-05,
+      "loss": 1.3903,
+      "step": 15530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2559862713961632e-05,
+      "loss": 1.4021,
+      "step": 15535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.2502760956976877e-05,
+      "loss": 1.4184,
+      "step": 15540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.2445722394172973e-05,
+      "loss": 1.3844,
+      "step": 15545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2388747072061335e-05,
+      "loss": 1.3948,
+      "step": 15550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2331835037101823e-05,
+      "loss": 1.411,
+      "step": 15555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.2274986335702597e-05,
+      "loss": 1.4426,
+      "step": 15560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2218201014220263e-05,
+      "loss": 1.3654,
+      "step": 15565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2161479118959737e-05,
+      "loss": 1.414,
+      "step": 15570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2104820696174235e-05,
+      "loss": 1.3961,
+      "step": 15575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.204822579206509e-05,
+      "loss": 1.4196,
+      "step": 15580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.1991694452781975e-05,
+      "loss": 1.4202,
+      "step": 15585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.1935226724422686e-05,
+      "loss": 1.4266,
+      "step": 15590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.187882265303317e-05,
+      "loss": 1.4199,
+      "step": 15595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.182248228460738e-05,
+      "loss": 1.4131,
+      "step": 15600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1766205665087426e-05,
+      "loss": 1.4076,
+      "step": 15605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.170999284036338e-05,
+      "loss": 1.3784,
+      "step": 15610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.1653843856273325e-05,
+      "loss": 1.4066,
+      "step": 15615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.1597758758603236e-05,
+      "loss": 1.4271,
+      "step": 15620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.154173759308703e-05,
+      "loss": 1.3941,
+      "step": 15625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.1485780405406498e-05,
+      "loss": 1.377,
+      "step": 15630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.142988724119127e-05,
+      "loss": 1.4119,
+      "step": 15635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1374058146018693e-05,
+      "loss": 1.4651,
+      "step": 15640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.131829316541395e-05,
+      "loss": 1.3771,
+      "step": 15645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.126259234484992e-05,
+      "loss": 1.4179,
+      "step": 15650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.120695572974718e-05,
+      "loss": 1.3984,
+      "step": 15655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1151383365473875e-05,
+      "loss": 1.4018,
+      "step": 15660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.109587529734586e-05,
+      "loss": 1.4305,
+      "step": 15665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.1040431570626483e-05,
+      "loss": 1.4443,
+      "step": 15670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.0985052230526714e-05,
+      "loss": 1.423,
+      "step": 15675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.092973732220489e-05,
+      "loss": 1.3989,
+      "step": 15680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.0874486890766908e-05,
+      "loss": 1.3782,
+      "step": 15685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.0819300981266066e-05,
+      "loss": 1.3892,
+      "step": 15690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.0764179638703076e-05,
+      "loss": 1.3963,
+      "step": 15695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.070912290802589e-05,
+      "loss": 1.415,
+      "step": 15700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.0654130834129903e-05,
+      "loss": 1.4013,
+      "step": 15705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.0599203461857707e-05,
+      "loss": 1.4032,
+      "step": 15710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.054434083599921e-05,
+      "loss": 1.4276,
+      "step": 15715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.0489543001291402e-05,
+      "loss": 1.4058,
+      "step": 15720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.0434810002418547e-05,
+      "loss": 1.414,
+      "step": 15725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.0380141884012004e-05,
+      "loss": 1.3732,
+      "step": 15730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.0325538690650236e-05,
+      "loss": 1.3867,
+      "step": 15735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0271000466858726e-05,
+      "loss": 1.4122,
+      "step": 15740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.0216527257110006e-05,
+      "loss": 1.4012,
+      "step": 15745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.0162119105823607e-05,
+      "loss": 1.4146,
+      "step": 15750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.010777605736599e-05,
+      "loss": 1.3965,
+      "step": 15755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.0053498156050555e-05,
+      "loss": 1.3938,
+      "step": 15760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9999285446137518e-05,
+      "loss": 1.3813,
+      "step": 15765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9945137971833983e-05,
+      "loss": 1.3778,
+      "step": 15770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9891055777293865e-05,
+      "loss": 1.4193,
+      "step": 15775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9837038906617843e-05,
+      "loss": 1.4143,
+      "step": 15780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9783087403853273e-05,
+      "loss": 1.4312,
+      "step": 15785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9729201312994273e-05,
+      "loss": 1.4178,
+      "step": 15790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9675380677981603e-05,
+      "loss": 1.4112,
+      "step": 15795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.962162554270267e-05,
+      "loss": 1.3914,
+      "step": 15800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.95679359509914e-05,
+      "loss": 1.4105,
+      "step": 15805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.951431194662834e-05,
+      "loss": 1.3702,
+      "step": 15810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.946075357334053e-05,
+      "loss": 1.3917,
+      "step": 15815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9407260874801513e-05,
+      "loss": 1.3756,
+      "step": 15820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9353833894631247e-05,
+      "loss": 1.3712,
+      "step": 15825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9300472676396076e-05,
+      "loss": 1.3687,
+      "step": 15830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9247177263608794e-05,
+      "loss": 1.4126,
+      "step": 15835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9193947699728488e-05,
+      "loss": 1.4815,
+      "step": 15840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9140784028160574e-05,
+      "loss": 1.3857,
+      "step": 15845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.908768629225669e-05,
+      "loss": 1.4191,
+      "step": 15850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9034654535314767e-05,
+      "loss": 1.3807,
+      "step": 15855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8981688800578877e-05,
+      "loss": 1.454,
+      "step": 15860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8928789131239343e-05,
+      "loss": 1.3978,
+      "step": 15865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.887595557043248e-05,
+      "loss": 1.3991,
+      "step": 15870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8823188161240813e-05,
+      "loss": 1.4167,
+      "step": 15875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8770486946692876e-05,
+      "loss": 1.4233,
+      "step": 15880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.8717851969763266e-05,
+      "loss": 1.4152,
+      "step": 15885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.866528327337249e-05,
+      "loss": 1.4035,
+      "step": 15890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.861278090038705e-05,
+      "loss": 1.3953,
+      "step": 15895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8560344893619396e-05,
+      "loss": 1.4082,
+      "step": 15900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.850797529582785e-05,
+      "loss": 1.4083,
+      "step": 15905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.8455672149716496e-05,
+      "loss": 1.4182,
+      "step": 15910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.840343549793535e-05,
+      "loss": 1.3979,
+      "step": 15915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.835126538308013e-05,
+      "loss": 1.4196,
+      "step": 15920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8299161847692358e-05,
+      "loss": 1.3812,
+      "step": 15925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.8247124934259186e-05,
+      "loss": 1.4142,
+      "step": 15930
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.819515468521349e-05,
+      "loss": 1.4185,
+      "step": 15935
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8143251142933793e-05,
+      "loss": 1.4173,
+      "step": 15940
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.809141434974423e-05,
+      "loss": 1.4131,
+      "step": 15945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.803964434791442e-05,
+      "loss": 1.4071,
+      "step": 15950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.7987941179659608e-05,
+      "loss": 1.356,
+      "step": 15955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.793630488714053e-05,
+      "loss": 1.4231,
+      "step": 15960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.788473551246339e-05,
+      "loss": 1.3839,
+      "step": 15965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7833233097679746e-05,
+      "loss": 1.4259,
+      "step": 15970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.778179768478666e-05,
+      "loss": 1.4312,
+      "step": 15975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.7730429315726494e-05,
+      "loss": 1.3773,
+      "step": 15980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7679128032387004e-05,
+      "loss": 1.3844,
+      "step": 15985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.762789387660113e-05,
+      "loss": 1.4109,
+      "step": 15990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7576726890147177e-05,
+      "loss": 1.3568,
+      "step": 15995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7525627114748645e-05,
+      "loss": 1.4048,
+      "step": 16000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7474594592074235e-05,
+      "loss": 1.4195,
+      "step": 16005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.742362936373776e-05,
+      "loss": 1.4214,
+      "step": 16010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.737273147129821e-05,
+      "loss": 1.41,
+      "step": 16015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7321900956259653e-05,
+      "loss": 1.4071,
+      "step": 16020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.727113786007125e-05,
+      "loss": 1.3967,
+      "step": 16025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7220442224127097e-05,
+      "loss": 1.3973,
+      "step": 16030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7169814089766344e-05,
+      "loss": 1.4093,
+      "step": 16035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7119253498273113e-05,
+      "loss": 1.4139,
+      "step": 16040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7068760490876422e-05,
+      "loss": 1.3937,
+      "step": 16045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.701833510875015e-05,
+      "loss": 1.4067,
+      "step": 16050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.696797739301308e-05,
+      "loss": 1.3547,
+      "step": 16055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6917687384728785e-05,
+      "loss": 1.4478,
+      "step": 16060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.686746512490569e-05,
+      "loss": 1.4055,
+      "step": 16065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6817310654496852e-05,
+      "loss": 1.4043,
+      "step": 16070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6767224014400173e-05,
+      "loss": 1.4098,
+      "step": 16075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6717205245458178e-05,
+      "loss": 1.4182,
+      "step": 16080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6667254388458088e-05,
+      "loss": 1.3897,
+      "step": 16085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.661737148413167e-05,
+      "loss": 1.3928,
+      "step": 16090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6567556573155374e-05,
+      "loss": 1.401,
+      "step": 16095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.6517809696150143e-05,
+      "loss": 1.4434,
+      "step": 16100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.64681308936815e-05,
+      "loss": 1.4293,
+      "step": 16105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.641852020625937e-05,
+      "loss": 1.403,
+      "step": 16110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.6368977674338216e-05,
+      "loss": 1.4235,
+      "step": 16115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.631950333831688e-05,
+      "loss": 1.4061,
+      "step": 16120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6270097238538597e-05,
+      "loss": 1.4604,
+      "step": 16125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.6220759415290998e-05,
+      "loss": 1.396,
+      "step": 16130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6171489908805992e-05,
+      "loss": 1.3742,
+      "step": 16135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6122288759259795e-05,
+      "loss": 1.4133,
+      "step": 16140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6073156006772893e-05,
+      "loss": 1.3686,
+      "step": 16145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6024091691410013e-05,
+      "loss": 1.3744,
+      "step": 16150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.597509585318001e-05,
+      "loss": 1.3963,
+      "step": 16155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.592616853203597e-05,
+      "loss": 1.3879,
+      "step": 16160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.587730976787508e-05,
+      "loss": 1.4037,
+      "step": 16165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.582851960053865e-05,
+      "loss": 1.3981,
+      "step": 16170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.577979806981198e-05,
+      "loss": 1.4068,
+      "step": 16175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.573114521542447e-05,
+      "loss": 1.4194,
+      "step": 16180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5682561077049496e-05,
+      "loss": 1.4033,
+      "step": 16185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5634045694304412e-05,
+      "loss": 1.4413,
+      "step": 16190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5585599106750515e-05,
+      "loss": 1.3756,
+      "step": 16195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.553722135389294e-05,
+      "loss": 1.4271,
+      "step": 16200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.548891247518075e-05,
+      "loss": 1.4217,
+      "step": 16205
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5440672510006848e-05,
+      "loss": 1.418,
+      "step": 16210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.5392501497707945e-05,
+      "loss": 1.3715,
+      "step": 16215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5344399477564462e-05,
+      "loss": 1.3594,
+      "step": 16220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.529636648880063e-05,
+      "loss": 1.4132,
+      "step": 16225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5248402570584353e-05,
+      "loss": 1.4164,
+      "step": 16230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.520050776202726e-05,
+      "loss": 1.4124,
+      "step": 16235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5152682102184546e-05,
+      "loss": 1.3827,
+      "step": 16240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5104925630055078e-05,
+      "loss": 1.4168,
+      "step": 16245
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5057238384581296e-05,
+      "loss": 1.4161,
+      "step": 16250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5009620404649193e-05,
+      "loss": 1.424,
+      "step": 16255
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4962071729088255e-05,
+      "loss": 1.3947,
+      "step": 16260
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4914592396671468e-05,
+      "loss": 1.4185,
+      "step": 16265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.486718244611528e-05,
+      "loss": 1.4036,
+      "step": 16270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.481984191607959e-05,
+      "loss": 1.397,
+      "step": 16275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.477257084516761e-05,
+      "loss": 1.4109,
+      "step": 16280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4725369271925982e-05,
+      "loss": 1.3913,
+      "step": 16285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4678237234844649e-05,
+      "loss": 1.4272,
+      "step": 16290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4631174772356881e-05,
+      "loss": 1.4109,
+      "step": 16295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4584181922839157e-05,
+      "loss": 1.4192,
+      "step": 16300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.4537258724611235e-05,
+      "loss": 1.3831,
+      "step": 16305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4490405215936066e-05,
+      "loss": 1.3819,
+      "step": 16310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4443621435019793e-05,
+      "loss": 1.3904,
+      "step": 16315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4396907420011651e-05,
+      "loss": 1.4392,
+      "step": 16320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4350263209004034e-05,
+      "loss": 1.453,
+      "step": 16325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4303688840032381e-05,
+      "loss": 1.4595,
+      "step": 16330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4257184351075237e-05,
+      "loss": 1.4095,
+      "step": 16335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4210749780054066e-05,
+      "loss": 1.3962,
+      "step": 16340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4164385164833394e-05,
+      "loss": 1.3973,
+      "step": 16345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4118090543220697e-05,
+      "loss": 1.4014,
+      "step": 16350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.4071865952966368e-05,
+      "loss": 1.3986,
+      "step": 16355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4025711431763644e-05,
+      "loss": 1.3975,
+      "step": 16360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3979627017248687e-05,
+      "loss": 1.4011,
+      "step": 16365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.393361274700049e-05,
+      "loss": 1.4472,
+      "step": 16370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3887668658540842e-05,
+      "loss": 1.4033,
+      "step": 16375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3841794789334239e-05,
+      "loss": 1.4049,
+      "step": 16380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3795991176788004e-05,
+      "loss": 1.4162,
+      "step": 16385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3750257858252124e-05,
+      "loss": 1.406,
+      "step": 16390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3704594871019305e-05,
+      "loss": 1.425,
+      "step": 16395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3659002252324838e-05,
+      "loss": 1.3878,
+      "step": 16400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3613480039346682e-05,
+      "loss": 1.4231,
+      "step": 16405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3568028269205391e-05,
+      "loss": 1.4099,
+      "step": 16410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3522646978964027e-05,
+      "loss": 1.4498,
+      "step": 16415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3477336205628233e-05,
+      "loss": 1.435,
+      "step": 16420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.3432095986146109e-05,
+      "loss": 1.4224,
+      "step": 16425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3386926357408257e-05,
+      "loss": 1.3966,
+      "step": 16430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3341827356247682e-05,
+      "loss": 1.3741,
+      "step": 16435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3296799019439865e-05,
+      "loss": 1.4017,
+      "step": 16440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3251841383702557e-05,
+      "loss": 1.4242,
+      "step": 16445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3206954485695944e-05,
+      "loss": 1.3895,
+      "step": 16450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3162138362022491e-05,
+      "loss": 1.405,
+      "step": 16455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3117393049226978e-05,
+      "loss": 1.4151,
+      "step": 16460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3072718583796405e-05,
+      "loss": 1.3724,
+      "step": 16465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3028115002160035e-05,
+      "loss": 1.4355,
+      "step": 16470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2983582340689304e-05,
+      "loss": 1.4304,
+      "step": 16475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2939120635697855e-05,
+      "loss": 1.3647,
+      "step": 16480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2894729923441407e-05,
+      "loss": 1.3749,
+      "step": 16485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2850410240117849e-05,
+      "loss": 1.403,
+      "step": 16490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.280616162186713e-05,
+      "loss": 1.4169,
+      "step": 16495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2761984104771252e-05,
+      "loss": 1.3677,
+      "step": 16500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.271787772485421e-05,
+      "loss": 1.4086,
+      "step": 16505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2673842518082024e-05,
+      "loss": 1.4118,
+      "step": 16510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.262987852036267e-05,
+      "loss": 1.4067,
+      "step": 16515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2585985767546083e-05,
+      "loss": 1.4026,
+      "step": 16520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2542164295424031e-05,
+      "loss": 1.4008,
+      "step": 16525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.249841413973022e-05,
+      "loss": 1.3992,
+      "step": 16530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2454735336140167e-05,
+      "loss": 1.3753,
+      "step": 16535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2411127920271271e-05,
+      "loss": 1.4067,
+      "step": 16540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2367591927682598e-05,
+      "loss": 1.4082,
+      "step": 16545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2324127393875084e-05,
+      "loss": 1.431,
+      "step": 16550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2280734354291346e-05,
+      "loss": 1.3781,
+      "step": 16555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2237412844315722e-05,
+      "loss": 1.4195,
+      "step": 16560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2194162899274208e-05,
+      "loss": 1.3911,
+      "step": 16565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.215098455443443e-05,
+      "loss": 1.3816,
+      "step": 16570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2107877845005644e-05,
+      "loss": 1.4202,
+      "step": 16575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2064842806138698e-05,
+      "loss": 1.3935,
+      "step": 16580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2021879472926023e-05,
+      "loss": 1.4332,
+      "step": 16585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1978987880401493e-05,
+      "loss": 1.3974,
+      "step": 16590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.1936168063540554e-05,
+      "loss": 1.418,
+      "step": 16595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1893420057260118e-05,
+      "loss": 1.4039,
+      "step": 16600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1850743896418537e-05,
+      "loss": 1.3886,
+      "step": 16605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1808139615815527e-05,
+      "loss": 1.3926,
+      "step": 16610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1765607250192245e-05,
+      "loss": 1.4107,
+      "step": 16615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1723146834231214e-05,
+      "loss": 1.4003,
+      "step": 16620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1680758402556257e-05,
+      "loss": 1.4199,
+      "step": 16625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1638441989732473e-05,
+      "loss": 1.3873,
+      "step": 16630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1596197630266292e-05,
+      "loss": 1.4124,
+      "step": 16635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1554025358605369e-05,
+      "loss": 1.4054,
+      "step": 16640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1511925209138575e-05,
+      "loss": 1.4418,
+      "step": 16645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.1469897216195924e-05,
+      "loss": 1.4035,
+      "step": 16650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.142794141404866e-05,
+      "loss": 1.3955,
+      "step": 16655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1386057836909137e-05,
+      "loss": 1.3921,
+      "step": 16660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1344246518930823e-05,
+      "loss": 1.4074,
+      "step": 16665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1302507494208191e-05,
+      "loss": 1.3626,
+      "step": 16670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1260840796776873e-05,
+      "loss": 1.4112,
+      "step": 16675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1219246460613452e-05,
+      "loss": 1.4386,
+      "step": 16680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1177724519635547e-05,
+      "loss": 1.4362,
+      "step": 16685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.113627500770167e-05,
+      "loss": 1.3881,
+      "step": 16690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.109489795861135e-05,
+      "loss": 1.392,
+      "step": 16695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1053593406105001e-05,
+      "loss": 1.3767,
+      "step": 16700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1012361383863946e-05,
+      "loss": 1.3679,
+      "step": 16705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0971201925510288e-05,
+      "loss": 1.4073,
+      "step": 16710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0930115064607016e-05,
+      "loss": 1.4165,
+      "step": 16715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0889100834657917e-05,
+      "loss": 1.403,
+      "step": 16720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0848159269107538e-05,
+      "loss": 1.3617,
+      "step": 16725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0807290401341219e-05,
+      "loss": 1.3885,
+      "step": 16730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0766494264684934e-05,
+      "loss": 1.3835,
+      "step": 16735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0725770892405407e-05,
+      "loss": 1.3934,
+      "step": 16740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0685120317710029e-05,
+      "loss": 1.3962,
+      "step": 16745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.064454257374683e-05,
+      "loss": 1.4156,
+      "step": 16750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0604037693604396e-05,
+      "loss": 1.427,
+      "step": 16755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0563605710311974e-05,
+      "loss": 1.3918,
+      "step": 16760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0523246656839314e-05,
+      "loss": 1.4136,
+      "step": 16765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0482960566096733e-05,
+      "loss": 1.4192,
+      "step": 16770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0442747470935022e-05,
+      "loss": 1.4162,
+      "step": 16775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0402607404145449e-05,
+      "loss": 1.3932,
+      "step": 16780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0362540398459752e-05,
+      "loss": 1.3889,
+      "step": 16785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0322546486550112e-05,
+      "loss": 1.4413,
+      "step": 16790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0282625701029037e-05,
+      "loss": 1.4081,
+      "step": 16795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0242778074449455e-05,
+      "loss": 1.4092,
+      "step": 16800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0203003639304643e-05,
+      "loss": 1.4168,
+      "step": 16805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0163302428028188e-05,
+      "loss": 1.4001,
+      "step": 16810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0123674472993916e-05,
+      "loss": 1.4336,
+      "step": 16815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0084119806516001e-05,
+      "loss": 1.4195,
+      "step": 16820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0044638460848798e-05,
+      "loss": 1.4218,
+      "step": 16825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.000523046818691e-05,
+      "loss": 1.3904,
+      "step": 16830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.965895860665075e-06,
+      "loss": 1.4194,
+      "step": 16835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.926634670358236e-06,
+      "loss": 1.386,
+      "step": 16840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.887446929281453e-06,
+      "loss": 1.4482,
+      "step": 16845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.848332669389916e-06,
+      "loss": 1.405,
+      "step": 16850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.809291922578823e-06,
+      "loss": 1.4228,
+      "step": 16855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.7703247206835e-06,
+      "loss": 1.3764,
+      "step": 16860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.731431095479281e-06,
+      "loss": 1.3871,
+      "step": 16865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.692611078681513e-06,
+      "loss": 1.4231,
+      "step": 16870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.653864701945469e-06,
+      "loss": 1.4045,
+      "step": 16875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.615191996866446e-06,
+      "loss": 1.4119,
+      "step": 16880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.576592994979617e-06,
+      "loss": 1.3788,
+      "step": 16885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.53806772776008e-06,
+      "loss": 1.4353,
+      "step": 16890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.499616226622766e-06,
+      "loss": 1.4092,
+      "step": 16895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.4612385229225e-06,
+      "loss": 1.4103,
+      "step": 16900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.422934647953929e-06,
+      "loss": 1.4503,
+      "step": 16905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.38470463295148e-06,
+      "loss": 1.4105,
+      "step": 16910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.346548509089326e-06,
+      "loss": 1.4304,
+      "step": 16915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.308466307481423e-06,
+      "loss": 1.3896,
+      "step": 16920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.270458059181452e-06,
+      "loss": 1.4248,
+      "step": 16925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.23252379518279e-06,
+      "loss": 1.4388,
+      "step": 16930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.194663546418436e-06,
+      "loss": 1.3883,
+      "step": 16935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.156877343761094e-06,
+      "loss": 1.3937,
+      "step": 16940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.11916521802304e-06,
+      "loss": 1.4132,
+      "step": 16945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.081527199956196e-06,
+      "loss": 1.4238,
+      "step": 16950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.043963320252025e-06,
+      "loss": 1.3871,
+      "step": 16955
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.006473609541511e-06,
+      "loss": 1.4014,
+      "step": 16960
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.96905809839519e-06,
+      "loss": 1.4254,
+      "step": 16965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.931716817323099e-06,
+      "loss": 1.4089,
+      "step": 16970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.89444979677474e-06,
+      "loss": 1.4142,
+      "step": 16975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.857257067139013e-06,
+      "loss": 1.4016,
+      "step": 16980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.820138658744304e-06,
+      "loss": 1.4131,
+      "step": 16985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.783094601858355e-06,
+      "loss": 1.4196,
+      "step": 16990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.746124926688325e-06,
+      "loss": 1.3525,
+      "step": 16995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.709229663380658e-06,
+      "loss": 1.3739,
+      "step": 17000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.67240884202113e-06,
+      "loss": 1.4177,
+      "step": 17005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.635662492634855e-06,
+      "loss": 1.3755,
+      "step": 17010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.59899064518619e-06,
+      "loss": 1.3941,
+      "step": 17015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.562393329578767e-06,
+      "loss": 1.4086,
+      "step": 17020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.525870575655392e-06,
+      "loss": 1.4255,
+      "step": 17025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.489422413198112e-06,
+      "loss": 1.3802,
+      "step": 17030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.453048871928138e-06,
+      "loss": 1.4057,
+      "step": 17035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.416749981505856e-06,
+      "loss": 1.3872,
+      "step": 17040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.380525771530701e-06,
+      "loss": 1.4199,
+      "step": 17045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.34437627154131e-06,
+      "loss": 1.3667,
+      "step": 17050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.308301511015327e-06,
+      "loss": 1.4286,
+      "step": 17055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.272301519369519e-06,
+      "loss": 1.4153,
+      "step": 17060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.236376325959583e-06,
+      "loss": 1.4186,
+      "step": 17065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.200525960080308e-06,
+      "loss": 1.387,
+      "step": 17070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.16475045096543e-06,
+      "loss": 1.418,
+      "step": 17075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.129049827787693e-06,
+      "loss": 1.3995,
+      "step": 17080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.093424119658678e-06,
+      "loss": 1.3728,
+      "step": 17085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.057873355628964e-06,
+      "loss": 1.3687,
+      "step": 17090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.022397564687989e-06,
+      "loss": 1.4024,
+      "step": 17095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.986996775764077e-06,
+      "loss": 1.4226,
+      "step": 17100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.951671017724316e-06,
+      "loss": 1.4087,
+      "step": 17105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.916420319374707e-06,
+      "loss": 1.3865,
+      "step": 17110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.88124470945999e-06,
+      "loss": 1.413,
+      "step": 17115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.846144216663697e-06,
+      "loss": 1.3887,
+      "step": 17120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.811118869608081e-06,
+      "loss": 1.4014,
+      "step": 17125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.776168696854147e-06,
+      "loss": 1.4044,
+      "step": 17130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.741293726901589e-06,
+      "loss": 1.4081,
+      "step": 17135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.70649398818879e-06,
+      "loss": 1.3762,
+      "step": 17140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.671769509092741e-06,
+      "loss": 1.3945,
+      "step": 17145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.637120317929114e-06,
+      "loss": 1.374,
+      "step": 17150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.6025464429521635e-06,
+      "loss": 1.3886,
+      "step": 17155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.56804791235477e-06,
+      "loss": 1.3781,
+      "step": 17160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.533624754268287e-06,
+      "loss": 1.4105,
+      "step": 17165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.499276996762694e-06,
+      "loss": 1.4119,
+      "step": 17170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.465004667846431e-06,
+      "loss": 1.4016,
+      "step": 17175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.430807795466488e-06,
+      "loss": 1.4167,
+      "step": 17180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.396686407508246e-06,
+      "loss": 1.4252,
+      "step": 17185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.362640531795606e-06,
+      "loss": 1.3954,
+      "step": 17190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.328670196090836e-06,
+      "loss": 1.3732,
+      "step": 17195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.294775428094669e-06,
+      "loss": 1.3902,
+      "step": 17200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.26095625544615e-06,
+      "loss": 1.39,
+      "step": 17205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.227212705722719e-06,
+      "loss": 1.444,
+      "step": 17210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.1935448064401445e-06,
+      "loss": 1.4021,
+      "step": 17215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.159952585052532e-06,
+      "loss": 1.4215,
+      "step": 17220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.126436068952202e-06,
+      "loss": 1.4201,
+      "step": 17225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.092995285469816e-06,
+      "loss": 1.4256,
+      "step": 17230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.05963026187425e-06,
+      "loss": 1.4086,
+      "step": 17235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.026341025372629e-06,
+      "loss": 1.3837,
+      "step": 17240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.993127603110216e-06,
+      "loss": 1.4214,
+      "step": 17245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.959990022170515e-06,
+      "loss": 1.4172,
+      "step": 17250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.926928309575154e-06,
+      "loss": 1.4337,
+      "step": 17255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.893942492283934e-06,
+      "loss": 1.4106,
+      "step": 17260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.861032597194683e-06,
+      "loss": 1.3771,
+      "step": 17265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.828198651143425e-06,
+      "loss": 1.4111,
+      "step": 17270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.795440680904164e-06,
+      "loss": 1.3926,
+      "step": 17275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.762758713189044e-06,
+      "loss": 1.4102,
+      "step": 17280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.730152774648113e-06,
+      "loss": 1.3879,
+      "step": 17285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.697622891869515e-06,
+      "loss": 1.3946,
+      "step": 17290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.665169091379364e-06,
+      "loss": 1.3882,
+      "step": 17295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.63279139964168e-06,
+      "loss": 1.3692,
+      "step": 17300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.6004898430585e-06,
+      "loss": 1.3792,
+      "step": 17305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.568264447969697e-06,
+      "loss": 1.4321,
+      "step": 17310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.536115240653096e-06,
+      "loss": 1.445,
+      "step": 17315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.504042247324371e-06,
+      "loss": 1.411,
+      "step": 17320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.47204549413708e-06,
+      "loss": 1.4009,
+      "step": 17325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.440125007182551e-06,
+      "loss": 1.4019,
+      "step": 17330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.408280812489964e-06,
+      "loss": 1.4255,
+      "step": 17335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.37651293602628e-06,
+      "loss": 1.3937,
+      "step": 17340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.344821403696255e-06,
+      "loss": 1.3765,
+      "step": 17345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.313206241342328e-06,
+      "loss": 1.3954,
+      "step": 17350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.281667474744712e-06,
+      "loss": 1.4106,
+      "step": 17355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.2502051296213226e-06,
+      "loss": 1.4296,
+      "step": 17360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.2188192316277374e-06,
+      "loss": 1.3906,
+      "step": 17365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.1875098063571835e-06,
+      "loss": 1.4083,
+      "step": 17370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.156276879340583e-06,
+      "loss": 1.4249,
+      "step": 17375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.125120476046431e-06,
+      "loss": 1.396,
+      "step": 17380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.094040621880837e-06,
+      "loss": 1.3768,
+      "step": 17385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.0630373421875055e-06,
+      "loss": 1.4468,
+      "step": 17390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.032110662247659e-06,
+      "loss": 1.4076,
+      "step": 17395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.0012606072800905e-06,
+      "loss": 1.387,
+      "step": 17400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.970487202441122e-06,
+      "loss": 1.3739,
+      "step": 17405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.939790472824535e-06,
+      "loss": 1.4175,
+      "step": 17410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.909170443461598e-06,
+      "loss": 1.421,
+      "step": 17415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.878627139321047e-06,
+      "loss": 1.4289,
+      "step": 17420
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.848160585309048e-06,
+      "loss": 1.408,
+      "step": 17425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.817770806269207e-06,
+      "loss": 1.4263,
+      "step": 17430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.787457826982457e-06,
+      "loss": 1.4055,
+      "step": 17435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.757221672167168e-06,
+      "loss": 1.4094,
+      "step": 17440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.727062366479041e-06,
+      "loss": 1.3952,
+      "step": 17445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.696979934511137e-06,
+      "loss": 1.3921,
+      "step": 17450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.666974400793779e-06,
+      "loss": 1.4552,
+      "step": 17455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.637045789794626e-06,
+      "loss": 1.4358,
+      "step": 17460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.607194125918602e-06,
+      "loss": 1.43,
+      "step": 17465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.577419433507891e-06,
+      "loss": 1.4005,
+      "step": 17470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.547721736841871e-06,
+      "loss": 1.4112,
+      "step": 17475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.518101060137204e-06,
+      "loss": 1.3764,
+      "step": 17480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.488557427547692e-06,
+      "loss": 1.4062,
+      "step": 17485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.459090863164351e-06,
+      "loss": 1.4151,
+      "step": 17490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.429701391015296e-06,
+      "loss": 1.4052,
+      "step": 17495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.400389035065845e-06,
+      "loss": 1.4168,
+      "step": 17500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.371153819218389e-06,
+      "loss": 1.4148,
+      "step": 17505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.341995767312435e-06,
+      "loss": 1.4043,
+      "step": 17510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.312914903124566e-06,
+      "loss": 1.398,
+      "step": 17515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.2839112503684e-06,
+      "loss": 1.426,
+      "step": 17520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.254984832694632e-06,
+      "loss": 1.3925,
+      "step": 17525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.226135673690957e-06,
+      "loss": 1.3968,
+      "step": 17530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.19736379688206e-06,
+      "loss": 1.3513,
+      "step": 17535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.168669225729616e-06,
+      "loss": 1.3746,
+      "step": 17540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.140051983632266e-06,
+      "loss": 1.4029,
+      "step": 17545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.111512093925619e-06,
+      "loss": 1.4421,
+      "step": 17550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.083049579882149e-06,
+      "loss": 1.3623,
+      "step": 17555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.054664464711267e-06,
+      "loss": 1.4099,
+      "step": 17560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.026356771559282e-06,
+      "loss": 1.4274,
+      "step": 17565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.998126523509361e-06,
+      "loss": 1.3686,
+      "step": 17570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.969973743581502e-06,
+      "loss": 1.3974,
+      "step": 17575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.941898454732563e-06,
+      "loss": 1.4286,
+      "step": 17580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.913900679856176e-06,
+      "loss": 1.4128,
+      "step": 17585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.885980441782823e-06,
+      "loss": 1.4221,
+      "step": 17590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.858137763279702e-06,
+      "loss": 1.403,
+      "step": 17595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.830372667050753e-06,
+      "loss": 1.4253,
+      "step": 17600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.802685175736732e-06,
+      "loss": 1.393,
+      "step": 17605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.775075311915045e-06,
+      "loss": 1.4162,
+      "step": 17610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.747543098099838e-06,
+      "loss": 1.3916,
+      "step": 17615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.720088556741897e-06,
+      "loss": 1.4114,
+      "step": 17620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.6927117102287034e-06,
+      "loss": 1.375,
+      "step": 17625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.665412580884365e-06,
+      "loss": 1.4022,
+      "step": 17630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.638191190969665e-06,
+      "loss": 1.4275,
+      "step": 17635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.611047562681903e-06,
+      "loss": 1.3914,
+      "step": 17640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.58398171815504e-06,
+      "loss": 1.3729,
+      "step": 17645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.556993679459587e-06,
+      "loss": 1.3844,
+      "step": 17650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.530083468602631e-06,
+      "loss": 1.4087,
+      "step": 17655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.503251107527751e-06,
+      "loss": 1.4423,
+      "step": 17660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.476496618115078e-06,
+      "loss": 1.3984,
+      "step": 17665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.449820022181239e-06,
+      "loss": 1.4143,
+      "step": 17670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.423221341479344e-06,
+      "loss": 1.3943,
+      "step": 17675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.396700597698955e-06,
+      "loss": 1.3885,
+      "step": 17680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.3702578124660834e-06,
+      "loss": 1.3702,
+      "step": 17685
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.34389300734318e-06,
+      "loss": 1.4119,
+      "step": 17690
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.317606203829127e-06,
+      "loss": 1.3952,
+      "step": 17695
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.291397423359156e-06,
+      "loss": 1.4155,
+      "step": 17700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.265266687304892e-06,
+      "loss": 1.4023,
+      "step": 17705
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.239214016974335e-06,
+      "loss": 1.3853,
+      "step": 17710
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.213239433611848e-06,
+      "loss": 1.4075,
+      "step": 17715
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.1873429583980325e-06,
+      "loss": 1.4139,
+      "step": 17720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.161524612449896e-06,
+      "loss": 1.4018,
+      "step": 17725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.135784416820665e-06,
+      "loss": 1.4292,
+      "step": 17730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.110122392499915e-06,
+      "loss": 1.3752,
+      "step": 17735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.0845385604133755e-06,
+      "loss": 1.3981,
+      "step": 17740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.059032941423113e-06,
+      "loss": 1.3676,
+      "step": 17745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.033605556327347e-06,
+      "loss": 1.3836,
+      "step": 17750
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.008256425860546e-06,
+      "loss": 1.4058,
+      "step": 17755
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.982985570693354e-06,
+      "loss": 1.4037,
+      "step": 17760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.957793011432564e-06,
+      "loss": 1.4318,
+      "step": 17765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.932678768621145e-06,
+      "loss": 1.381,
+      "step": 17770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.907642862738214e-06,
+      "loss": 1.4374,
+      "step": 17775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.882685314199009e-06,
+      "loss": 1.3676,
+      "step": 17780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.857806143354814e-06,
+      "loss": 1.4109,
+      "step": 17785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.833005370493081e-06,
+      "loss": 1.3809,
+      "step": 17790
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.808283015837277e-06,
+      "loss": 1.4036,
+      "step": 17795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.7836390995469873e-06,
+      "loss": 1.383,
+      "step": 17800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.7590736417177365e-06,
+      "loss": 1.4214,
+      "step": 17805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.7345866623811677e-06,
+      "loss": 1.381,
+      "step": 17810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.7101781815048753e-06,
+      "loss": 1.3782,
+      "step": 17815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.6858482189924716e-06,
+      "loss": 1.4236,
+      "step": 17820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.6615967946835084e-06,
+      "loss": 1.3932,
+      "step": 17825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.637423928353523e-06,
+      "loss": 1.3883,
+      "step": 17830
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.6133296397139804e-06,
+      "loss": 1.4071,
+      "step": 17835
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.5893139484122982e-06,
+      "loss": 1.3812,
+      "step": 17840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.565376874031756e-06,
+      "loss": 1.3953,
+      "step": 17845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.541518436091562e-06,
+      "loss": 1.4084,
+      "step": 17850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.517738654046776e-06,
+      "loss": 1.4153,
+      "step": 17855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.4940375472883536e-06,
+      "loss": 1.4134,
+      "step": 17860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.470415135143046e-06,
+      "loss": 1.4192,
+      "step": 17865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.446871436873478e-06,
+      "loss": 1.4571,
+      "step": 17870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.42340647167807e-06,
+      "loss": 1.3937,
+      "step": 17875
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.400020258691061e-06,
+      "loss": 1.3724,
+      "step": 17880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.3767128169824304e-06,
+      "loss": 1.4317,
+      "step": 17885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.353484165557941e-06,
+      "loss": 1.4134,
+      "step": 17890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.33033432335913e-06,
+      "loss": 1.4244,
+      "step": 17895
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.307263309263242e-06,
+      "loss": 1.3976,
+      "step": 17900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.284271142083284e-06,
+      "loss": 1.4189,
+      "step": 17905
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.2613578405679023e-06,
+      "loss": 1.3897,
+      "step": 17910
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.238523423401496e-06,
+      "loss": 1.4193,
+      "step": 17915
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.2157679092040927e-06,
+      "loss": 1.4181,
+      "step": 17920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.193091316531427e-06,
+      "loss": 1.407,
+      "step": 17925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.1704936638748296e-06,
+      "loss": 1.4167,
+      "step": 17930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.1479749696612713e-06,
+      "loss": 1.3896,
+      "step": 17935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.1255352522533755e-06,
+      "loss": 1.4004,
+      "step": 17940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.1031745299493266e-06,
+      "loss": 1.395,
+      "step": 17945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.0808928209828837e-06,
+      "loss": 1.408,
+      "step": 17950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.058690143523424e-06,
+      "loss": 1.3846,
+      "step": 17955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.0365665156758315e-06,
+      "loss": 1.4238,
+      "step": 17960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.014521955480565e-06,
+      "loss": 1.4062,
+      "step": 17965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9925564809135776e-06,
+      "loss": 1.3923,
+      "step": 17970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.970670109886353e-06,
+      "loss": 1.4395,
+      "step": 17975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.94886286024586e-06,
+      "loss": 1.4237,
+      "step": 17980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.927134749774585e-06,
+      "loss": 1.4126,
+      "step": 17985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5,
+      "learning_rate": 2.9054857961904216e-06,
+      "loss": 1.3602,
+      "step": 17990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.8839160171467485e-06,
+      "loss": 1.4304,
+      "step": 17995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.862425430232385e-06,
+      "loss": 1.4215,
+      "step": 18000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.84101405297158e-06,
+      "loss": 1.3593,
+      "step": 18005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8196819028239565e-06,
+      "loss": 1.4182,
+      "step": 18010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7984289971845657e-06,
+      "loss": 1.4155,
+      "step": 18015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.777255353383845e-06,
+      "loss": 1.3748,
+      "step": 18020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.756160988687573e-06,
+      "loss": 1.3857,
+      "step": 18025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.735145920296889e-06,
+      "loss": 1.4111,
+      "step": 18030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.7142101653482852e-06,
+      "loss": 1.4205,
+      "step": 18035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.693353740913562e-06,
+      "loss": 1.4234,
+      "step": 18040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6725766639998485e-06,
+      "loss": 1.4098,
+      "step": 18045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.651878951549536e-06,
+      "loss": 1.3929,
+      "step": 18050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.6312606204403343e-06,
+      "loss": 1.4238,
+      "step": 18055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.610721687485207e-06,
+      "loss": 1.3936,
+      "step": 18060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.5902621694324005e-06,
+      "loss": 1.3911,
+      "step": 18065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.569882082965358e-06,
+      "loss": 1.4094,
+      "step": 18070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.5495814447027643e-06,
+      "loss": 1.3685,
+      "step": 18075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.5293602711985444e-06,
+      "loss": 1.3835,
+      "step": 18080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.5092185789418078e-06,
+      "loss": 1.3872,
+      "step": 18085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.48915638435685e-06,
+      "loss": 1.3883,
+      "step": 18090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.469173703803129e-06,
+      "loss": 1.4209,
+      "step": 18095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4492705535753003e-06,
+      "loss": 1.389,
+      "step": 18100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.429446949903147e-06,
+      "loss": 1.4392,
+      "step": 18105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.409702908951561e-06,
+      "loss": 1.3789,
+      "step": 18110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.3900384468205974e-06,
+      "loss": 1.3633,
+      "step": 18115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3704535795454065e-06,
+      "loss": 1.3901,
+      "step": 18120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.350948323096214e-06,
+      "loss": 1.4299,
+      "step": 18125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.331522693378374e-06,
+      "loss": 1.3986,
+      "step": 18130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3121767062322387e-06,
+      "loss": 1.4388,
+      "step": 18135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.2929103774332882e-06,
+      "loss": 1.3955,
+      "step": 18140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2737237226920003e-06,
+      "loss": 1.4058,
+      "step": 18145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.2546167576539155e-06,
+      "loss": 1.4167,
+      "step": 18150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.2355894978995593e-06,
+      "loss": 1.4242,
+      "step": 18155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2166419589444875e-06,
+      "loss": 1.4274,
+      "step": 18160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1977741562392294e-06,
+      "loss": 1.414,
+      "step": 18165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.178986105169334e-06,
+      "loss": 1.4187,
+      "step": 18170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1602778210552564e-06,
+      "loss": 1.3972,
+      "step": 18175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.141649319152461e-06,
+      "loss": 1.3914,
+      "step": 18180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.1231006146513187e-06,
+      "loss": 1.4213,
+      "step": 18185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.1046317226771417e-06,
+      "loss": 1.4014,
+      "step": 18190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.086242658290194e-06,
+      "loss": 1.4132,
+      "step": 18195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.0679334364855806e-06,
+      "loss": 1.3834,
+      "step": 18200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.049704072193337e-06,
+      "loss": 1.4112,
+      "step": 18205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.031554580278394e-06,
+      "loss": 1.4272,
+      "step": 18210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.013484975540536e-06,
+      "loss": 1.4039,
+      "step": 18215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.995495272714376e-06,
+      "loss": 1.406,
+      "step": 18220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9775854864694134e-06,
+      "loss": 1.3711,
+      "step": 18225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.959755631409976e-06,
+      "loss": 1.4097,
+      "step": 18230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9420057220751907e-06,
+      "loss": 1.4077,
+      "step": 18235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.924335772939012e-06,
+      "loss": 1.3996,
+      "step": 18240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.90674579841017e-06,
+      "loss": 1.4285,
+      "step": 18245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8892358128322018e-06,
+      "loss": 1.43,
+      "step": 18250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8718058304834307e-06,
+      "loss": 1.4526,
+      "step": 18255
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8544558655768983e-06,
+      "loss": 1.4166,
+      "step": 18260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8371859322604323e-06,
+      "loss": 1.4201,
+      "step": 18265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8199960446165898e-06,
+      "loss": 1.3988,
+      "step": 18270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8028862166626691e-06,
+      "loss": 1.3912,
+      "step": 18275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7858564623506547e-06,
+      "loss": 1.3984,
+      "step": 18280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.768906795567249e-06,
+      "loss": 1.4118,
+      "step": 18285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7520372301338516e-06,
+      "loss": 1.37,
+      "step": 18290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7352477798065703e-06,
+      "loss": 1.3932,
+      "step": 18295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.718538458276131e-06,
+      "loss": 1.4033,
+      "step": 18300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.701909279167946e-06,
+      "loss": 1.402,
+      "step": 18305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6853602560421012e-06,
+      "loss": 1.4002,
+      "step": 18310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6688914023932801e-06,
+      "loss": 1.4128,
+      "step": 18315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6525027316507957e-06,
+      "loss": 1.3474,
+      "step": 18320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.6361942571786138e-06,
+      "loss": 1.3617,
+      "step": 18325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.619965992275274e-06,
+      "loss": 1.4078,
+      "step": 18330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.6038179501739138e-06,
+      "loss": 1.4057,
+      "step": 18335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.587750144042266e-06,
+      "loss": 1.3938,
+      "step": 18340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5717625869826168e-06,
+      "loss": 1.3701,
+      "step": 18345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.555855292031827e-06,
+      "loss": 1.4177,
+      "step": 18350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.540028272161309e-06,
+      "loss": 1.4019,
+      "step": 18355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5242815402770172e-06,
+      "loss": 1.3823,
+      "step": 18360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5086151092194356e-06,
+      "loss": 1.4226,
+      "step": 18365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4930289917635453e-06,
+      "loss": 1.3835,
+      "step": 18370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.4775232006188799e-06,
+      "loss": 1.3652,
+      "step": 18375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4620977484294362e-06,
+      "loss": 1.4092,
+      "step": 18380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4467526477737082e-06,
+      "loss": 1.4041,
+      "step": 18385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4314879111646861e-06,
+      "loss": 1.4034,
+      "step": 18390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4163035510498023e-06,
+      "loss": 1.4054,
+      "step": 18395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4011995798109522e-06,
+      "loss": 1.3963,
+      "step": 18400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.386176009764506e-06,
+      "loss": 1.4419,
+      "step": 18405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3712328531612306e-06,
+      "loss": 1.414,
+      "step": 18410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3563701221863567e-06,
+      "loss": 1.3657,
+      "step": 18415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3415878289595008e-06,
+      "loss": 1.4276,
+      "step": 18420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3268859855347093e-06,
+      "loss": 1.3805,
+      "step": 18425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.312264603900437e-06,
+      "loss": 1.4341,
+      "step": 18430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2977236959795025e-06,
+      "loss": 1.3819,
+      "step": 18435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2832632736290983e-06,
+      "loss": 1.4432,
+      "step": 18440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2688833486408257e-06,
+      "loss": 1.4064,
+      "step": 18445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.254583932740594e-06,
+      "loss": 1.4189,
+      "step": 18450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2403650375887088e-06,
+      "loss": 1.3865,
+      "step": 18455
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2262266747797847e-06,
+      "loss": 1.4215,
+      "step": 18460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2121688558427768e-06,
+      "loss": 1.3978,
+      "step": 18465
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1981915922409603e-06,
+      "loss": 1.4087,
+      "step": 18470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1842948953719403e-06,
+      "loss": 1.3789,
+      "step": 18475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1704787765675963e-06,
+      "loss": 1.3762,
+      "step": 18480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1567432470941163e-06,
+      "loss": 1.4136,
+      "step": 18485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1430883181519635e-06,
+      "loss": 1.4364,
+      "step": 18490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1295140008758864e-06,
+      "loss": 1.4176,
+      "step": 18495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1160203063349195e-06,
+      "loss": 1.4348,
+      "step": 18500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1026072455322945e-06,
+      "loss": 1.4194,
+      "step": 18505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.089274829405562e-06,
+      "loss": 1.3916,
+      "step": 18510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0760230688264593e-06,
+      "loss": 1.4221,
+      "step": 18515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0628519746009757e-06,
+      "loss": 1.3982,
+      "step": 18520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0497615574693309e-06,
+      "loss": 1.4214,
+      "step": 18525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0367518281059307e-06,
+      "loss": 1.4061,
+      "step": 18530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0238227971194004e-06,
+      "loss": 1.3746,
+      "step": 18535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.010974475052584e-06,
+      "loss": 1.4275,
+      "step": 18540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.982068723824677e-07,
+      "loss": 1.3601,
+      "step": 18545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.855199995202457e-07,
+      "loss": 1.3907,
+      "step": 18550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.729138668112648e-07,
+      "loss": 1.4076,
+      "step": 18555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.603884845350575e-07,
+      "loss": 1.4104,
+      "step": 18560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.479438629052873e-07,
+      "loss": 1.3915,
+      "step": 18565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.3558001206977e-07,
+      "loss": 1.4368,
+      "step": 18570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.232969421104521e-07,
+      "loss": 1.4154,
+      "step": 18575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.110946630434214e-07,
+      "loss": 1.4285,
+      "step": 18580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.989731848188743e-07,
+      "loss": 1.4052,
+      "step": 18585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.869325173211262e-07,
+      "loss": 1.3765,
+      "step": 18590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.749726703685901e-07,
+      "loss": 1.4041,
+      "step": 18595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.630936537137757e-07,
+      "loss": 1.3355,
+      "step": 18600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.51295477043279e-07,
+      "loss": 1.4103,
+      "step": 18605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.395781499777932e-07,
+      "loss": 1.3912,
+      "step": 18610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.279416820720531e-07,
+      "loss": 1.41,
+      "step": 18615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.163860828148906e-07,
+      "loss": 1.3804,
+      "step": 18620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.049113616291793e-07,
+      "loss": 1.4257,
+      "step": 18625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.935175278718232e-07,
+      "loss": 1.3851,
+      "step": 18630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.822045908337905e-07,
+      "loss": 1.3834,
+      "step": 18635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.709725597400908e-07,
+      "loss": 1.4218,
+      "step": 18640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.598214437497531e-07,
+      "loss": 1.3998,
+      "step": 18645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.487512519557815e-07,
+      "loss": 1.3893,
+      "step": 18650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.377619933852664e-07,
+      "loss": 1.3919,
+      "step": 18655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.268536769992507e-07,
+      "loss": 1.408,
+      "step": 18660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.160263116927968e-07,
+      "loss": 1.388,
+      "step": 18665
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.052799062949312e-07,
+      "loss": 1.4445,
+      "step": 18670
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.946144695686885e-07,
+      "loss": 1.3965,
+      "step": 18675
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.840300102110785e-07,
+      "loss": 1.4341,
+      "step": 18680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.735265368530641e-07,
+      "loss": 1.3783,
+      "step": 18685
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.631040580595605e-07,
+      "loss": 1.3876,
+      "step": 18690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.527625823294692e-07,
+      "loss": 1.3976,
+      "step": 18695
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.425021180956114e-07,
+      "loss": 1.3861,
+      "step": 18700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.323226737247723e-07,
+      "loss": 1.4201,
+      "step": 18705
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.222242575176341e-07,
+      "loss": 1.4344,
+      "step": 18710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.122068777088319e-07,
+      "loss": 1.433,
+      "step": 18715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.022705424669317e-07,
+      "loss": 1.4217,
+      "step": 18720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.924152598943966e-07,
+      "loss": 1.4262,
+      "step": 18725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.826410380275759e-07,
+      "loss": 1.3997,
+      "step": 18730
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.729478848367609e-07,
+      "loss": 1.4183,
+      "step": 18735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.633358082260954e-07,
+      "loss": 1.3599,
+      "step": 18740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.53804816033654e-07,
+      "loss": 1.408,
+      "step": 18745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.443549160313421e-07,
+      "loss": 1.404,
+      "step": 18750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.349861159249959e-07,
+      "loss": 1.4162,
+      "step": 18755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.256984233542595e-07,
+      "loss": 1.3979,
+      "step": 18760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.16491845892697e-07,
+      "loss": 1.3693,
+      "step": 18765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.073663910476811e-07,
+      "loss": 1.3507,
+      "step": 18770
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.983220662604482e-07,
+      "loss": 1.4163,
+      "step": 18775
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.893588789060988e-07,
+      "loss": 1.393,
+      "step": 18780
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.80476836293542e-07,
+      "loss": 1.4092,
+      "step": 18785
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.7167594566555064e-07,
+      "loss": 1.3873,
+      "step": 18790
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.6295621419868427e-07,
+      "loss": 1.401,
+      "step": 18795
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.5431764900334404e-07,
+      "loss": 1.4024,
+      "step": 18800
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.457602571237507e-07,
+      "loss": 1.3943,
+      "step": 18805
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.3728404553793344e-07,
+      "loss": 1.395,
+      "step": 18810
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.288890211576857e-07,
+      "loss": 1.3996,
+      "step": 18815
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.205751908286537e-07,
+      "loss": 1.4113,
+      "step": 18820
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.1234256133024785e-07,
+      "loss": 1.4166,
+      "step": 18825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.0419113937566475e-07,
+      "loss": 1.4103,
+      "step": 18830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.961209316118653e-07,
+      "loss": 1.3753,
+      "step": 18835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.8813194461961856e-07,
+      "loss": 1.4105,
+      "step": 18840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.8022418491344693e-07,
+      "loss": 1.4321,
+      "step": 18845
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.723976589416256e-07,
+      "loss": 1.4311,
+      "step": 18850
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.6465237308621615e-07,
+      "loss": 1.4195,
+      "step": 18855
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.5698833366299975e-07,
+      "loss": 1.4145,
+      "step": 18860
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.4940554692154405e-07,
+      "loss": 1.3957,
+      "step": 18865
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.41904019045125e-07,
+      "loss": 1.4162,
+      "step": 18870
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.344837561507719e-07,
+      "loss": 1.3873,
+      "step": 18875
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.2714476428925553e-07,
+      "loss": 1.4291,
+      "step": 18880
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.1988704944506677e-07,
+      "loss": 1.4272,
+      "step": 18885
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.127106175364158e-07,
+      "loss": 1.4175,
+      "step": 18890
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.05615474415244e-07,
+      "loss": 1.422,
+      "step": 18895
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.9860162586718974e-07,
+      "loss": 1.4079,
+      "step": 18900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.9166907761162264e-07,
+      "loss": 1.4483,
+      "step": 18905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.8481783530159843e-07,
+      "loss": 1.4063,
+      "step": 18910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.780479045238704e-07,
+      "loss": 1.4074,
+      "step": 18915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7135929079891156e-07,
+      "loss": 1.4116,
+      "step": 18920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6475199958085897e-07,
+      "loss": 1.3984,
+      "step": 18925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.582260362575584e-07,
+      "loss": 1.4304,
+      "step": 18930
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.5178140615051973e-07,
+      "loss": 1.404,
+      "step": 18935
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4541811451493925e-07,
+      "loss": 1.4332,
+      "step": 18940
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.391361665396885e-07,
+      "loss": 1.4137,
+      "step": 18945
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3293556734730326e-07,
+      "loss": 1.4516,
+      "step": 18950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.268163219939945e-07,
+      "loss": 1.4191,
+      "step": 18955
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.2077843546960408e-07,
+      "loss": 1.3984,
+      "step": 18960
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1482191269768247e-07,
+      "loss": 1.4172,
+      "step": 18965
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.089467585353777e-07,
+      "loss": 1.3938,
+      "step": 18970
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.0315297777353525e-07,
+      "loss": 1.4371,
+      "step": 18975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9744057513660928e-07,
+      "loss": 1.4102,
+      "step": 18980
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9180955528270706e-07,
+      "loss": 1.4085,
+      "step": 18985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8625992280357773e-07,
+      "loss": 1.3943,
+      "step": 18990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.8079168222461252e-07,
+      "loss": 1.4169,
+      "step": 18995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7540483800481122e-07,
+      "loss": 1.379,
+      "step": 19000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7009939453680456e-07,
+      "loss": 1.3613,
+      "step": 19005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.6487535614687633e-07,
+      "loss": 1.4412,
+      "step": 19010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5973272709487453e-07,
+      "loss": 1.4065,
+      "step": 19015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.5467151157431136e-07,
+      "loss": 1.3788,
+      "step": 19020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4969171371228552e-07,
+      "loss": 1.4031,
+      "step": 19025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.447933375695265e-07,
+      "loss": 1.376,
+      "step": 19030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3997638714033923e-07,
+      "loss": 1.3969,
+      "step": 19035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3524086635265942e-07,
+      "loss": 1.3911,
+      "step": 19040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.305867790679982e-07,
+      "loss": 1.4282,
+      "step": 19045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2601412908147536e-07,
+      "loss": 1.4233,
+      "step": 19050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2152292012181932e-07,
+      "loss": 1.4246,
+      "step": 19055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1711315585131166e-07,
+      "loss": 1.4038,
+      "step": 19060
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1278483986586486e-07,
+      "loss": 1.3726,
+      "step": 19065
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.085379756949223e-07,
+      "loss": 1.4142,
+      "step": 19070
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0437256680155827e-07,
+      "loss": 1.3689,
+      "step": 19075
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0028861658238909e-07,
+      "loss": 1.4363,
+      "step": 19080
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.628612836763973e-08,
+      "loss": 1.4179,
+      "step": 19085
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.236510542107723e-08,
+      "loss": 1.3981,
+      "step": 19090
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.85255509400662e-08,
+      "loss": 1.4081,
+      "step": 19095
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.476746805550218e-08,
+      "loss": 1.4231,
+      "step": 19100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.109085983188936e-08,
+      "loss": 1.3895,
+      "step": 19105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.74957292672629e-08,
+      "loss": 1.4086,
+      "step": 19110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.39820792932333e-08,
+      "loss": 1.3892,
+      "step": 19115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.054991277496425e-08,
+      "loss": 1.4197,
+      "step": 19120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.719923251116145e-08,
+      "loss": 1.4062,
+      "step": 19125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.393004123411706e-08,
+      "loss": 1.424,
+      "step": 19130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.074234160963199e-08,
+      "loss": 1.3763,
+      "step": 19135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.763613623709363e-08,
+      "loss": 1.4039,
+      "step": 19140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.461142764940919e-08,
+      "loss": 1.4028,
+      "step": 19145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.166821831305013e-08,
+      "loss": 1.3935,
+      "step": 19150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.880651062800779e-08,
+      "loss": 1.4228,
+      "step": 19155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.6026306927848814e-08,
+      "loss": 1.4005,
+      "step": 19160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.332760947962644e-08,
+      "loss": 1.4044,
+      "step": 19165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.0710420483980326e-08,
+      "loss": 1.4302,
+      "step": 19170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.817474207505889e-08,
+      "loss": 1.3764,
+      "step": 19175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.572057632055259e-08,
+      "loss": 1.4099,
+      "step": 19180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.3347925221682844e-08,
+      "loss": 1.3482,
+      "step": 19185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.1056790713202e-08,
+      "loss": 1.4299,
+      "step": 19190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.884717466338227e-08,
+      "loss": 1.4004,
+      "step": 19195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.6719078874026803e-08,
+      "loss": 1.4316,
+      "step": 19200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.4672505080458597e-08,
+      "loss": 1.3985,
+      "step": 19205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2707454951553797e-08,
+      "loss": 1.4001,
+      "step": 19210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.082393008966399e-08,
+      "loss": 1.4053,
+      "step": 19215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9021932030705015e-08,
+      "loss": 1.4332,
+      "step": 19220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7301462244079246e-08,
+      "loss": 1.4338,
+      "step": 19225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.5662522132742218e-08,
+      "loss": 1.3848,
+      "step": 19230
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4105113033124895e-08,
+      "loss": 1.4195,
+      "step": 19235
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2629236215211393e-08,
+      "loss": 1.3918,
+      "step": 19240
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1234892882494574e-08,
+      "loss": 1.4252,
+      "step": 19245
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.922084171953839e-09,
+      "loss": 1.4026,
+      "step": 19250
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.690811154121737e-09,
+      "loss": 1.3924,
+      "step": 19255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.541074833006257e-09,
+      "loss": 1.3749,
+      "step": 19260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.472876146168538e-09,
+      "loss": 1.4085,
+      "step": 19265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.486215964645158e-09,
+      "loss": 1.442,
+      "step": 19270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.581095092992538e-09,
+      "loss": 1.4407,
+      "step": 19275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.757514269286944e-09,
+      "loss": 1.4129,
+      "step": 19280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.0154741651022833e-09,
+      "loss": 1.4254,
+      "step": 19285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.354975385543412e-09,
+      "loss": 1.413,
+      "step": 19290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.776018469179519e-09,
+      "loss": 1.397,
+      "step": 19295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2786038881329455e-09,
+      "loss": 1.4139,
+      "step": 19300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.627320480125711e-10,
+      "loss": 1.4221,
+      "step": 19305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.284032879249168e-10,
+      "loss": 1.3758,
+      "step": 19310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.7561788049634827e-10,
+      "loss": 1.3815,
+      "step": 19315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0437603187307688e-10,
+      "loss": 1.4298,
+      "step": 19320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4677881676750105e-11,
+      "loss": 1.3775,
+      "step": 19325
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.4011043310165405,
+      "eval_runtime": 25105.0488,
+      "eval_samples_per_second": 6.821,
+      "eval_steps_per_second": 1.705,
+      "step": 19328
+    },
+    {
+      "epoch": 1.0,
+      "step": 19328,
+      "total_flos": 1.3590087469130318e+19,
+      "train_loss": 1.456397372168421,
+      "train_runtime": 104474.4259,
+      "train_samples_per_second": 1.48,
+      "train_steps_per_second": 0.185
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 19328,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.3590087469130318e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}